11.2.1 Legitimacy, Democracy, and Democratic Legitimacy

The concept of legitimacy, first, is commonly invoked as a frame for capturing the underlying normative value that democratic institutions should be designed to deliver. The etymological roots of the legitimacy concept can be traced back to the Latin root lex, meaning ‘law’, and derivatives such as legitimus, meaning ‘according to law’.Footnote ¹⁵ But in the early modern period the legitimacy idea evolved to take on more specialized political meanings: first capturing lawfully delegated political powers,Footnote ¹⁶ and hereditary rights of ruling monarchs;Footnote ¹⁷ and subsequently expanding to allow scope for subjecting even lawful power to critique in the name of some autonomous set of political agents or reasons.Footnote ¹⁸ While the precise sources of legitimacy’s distinctive political normativity remains a matter of ongoing dispute,Footnote ¹⁹ most theorists understand legitimacy standards as functioning in some sense to remedy political problems of normative dispute over the rightful institutional terms of political rule.Footnote ²⁰ Remedying such legitimacy problems is commonly valued for helping reconcile political rule with the normative judgements and commitments of free individuals,Footnote ²¹ and further for helping sustain beneficial political cooperation at low enforcement costs.Footnote ²²

This modern idea of legitimacy thus captures not any ideal institutional model of political rule, but rather a distinctive normative standard against which institutional models can be judged and designed. It is through this application that the idea of legitimacy has come to be commonly aligned with a reinvigorated modern interpretation of the old Greek idea of democracy – defined as a form of political rule or empowerment (arche) through which a political community or ‘people’ (demos) engages in collective self-governance. The compound idea of democratic legitimacy captures what results when democratic institutional instruments successfully function as political remedies to underlying legitimacy problems – resolving normative disputes over the institutional terms of political rule through the collectivization of power in democratic decision-making processes.

Not all contemporary justifications of democratic institutions appeal in this way to their value as instruments of legitimation: sometimes democracy is justified instead as a vehicle for some thicker moral conception of justice,Footnote ²³ or another instrumental moral purpose.Footnote ²⁴ Similarly, the concept of ‘legitimacy’ is sometimes appended to democracy in a more descriptive sense – denoting not its conformity to any substantive normative standard, but rather a generic attribution of political justifiability or acceptability, on whatever terms these may be understood.Footnote ²⁵ But it was as an instrument of legitimacy in its substantive normative sense that institutional ideals of democracy first gained widespread uptake in modern political thought and practice, under the influence of Jean-Jacques Rousseau’s proposed collectivist solution to the legitimacy problems of consolidating European States.Footnote ²⁶ And in this same Rousseauvian tradition, deliberative democratic models have attained contemporary theoretical support as instruments of normative legitimacy – entrenching the pairing of democracy with legitimacy as political solution and problem, as these function within the State-based institutional orders of modern political life.Footnote ²⁷

11.2.2 Democracy, Representation, and Transnational Legitimacy Problems

Alongside this normative idea of legitimacy, the concept of representation has had enormous theoretical influence in reshaping models of democracy for operation in the modern political world. Like the idea of democracy itself, the concept of representation marks out a model of the relationships and structures of political rule, without entailing any particular account of their (justifying or legitimizing) normative grounds. Most broadly, the concept of representation means ‘making present’ in a new sense something which is, in another sense, absent;Footnote ²⁸ as such, it has many meanings and applications outside the domain of political life. But when applied to the political context of democratic governance, the concept captures those political relationships and practices through which members of democratic constituencies can be made present in the activities of political rule – even when (due to constraints such as time, distance, resources, competence, or motivation) they do not participate in these directly.Footnote ²⁹

At the heart of all models of democratic representation is some kind of relationship between: the democratic constituencies who are (at least in some direct participatory sense) absent from activities of political rule, yet democratically entitled to be present; and their democratic representatives, whose characteristics and activities within sites and processes of political rule can make their represented constituencies in some sense present again. Competing theoretical models of democratic representation then vary in how they characterize these political agents and relationships, across several dimensions: the elements of constituencies’ attributes or activities that representative processes seek to make present; the elements of representatives’ attributes or activities through which this presence is fostered in sites and processes of political rule; and the features of the political relationships between representatives and constituents through which these representative practices are sustained.Footnote ³⁰

In designing and justifying models of political representation along these several dimensions, theorists have not typically undertaken systematic analysis of their implications for political problems of legitimacy, in the substantive normative sense of interest here. Rather, a standard analytical strategy has been to assess representative models against the intermediary concept of ‘democracy’ – assuming that whatever can be upheld as democratic warrants an accompanying attribution of normative legitimacy. For some representative theorists an assumption of such equivalence is merely implicit. Others, however, have argued explicitly that the normative assessment of representative institutional designs requires no appeal to any standards exogenous to the concept of democracy itself, but can rather be assessed simply for its ‘democraticity’ – meaning conformity with the conceptual constraints of the democratic idea.Footnote ³¹ More concretely, this has meant that the assessment of representative designs has mostly focused on the democratic function of representative inclusion – aimed at ensuring participants in political rule (arche) include all other members of their political community (demos) in their decision-making processes.Footnote ³² In this democratic interpretation of the representative challenge, normative evaluation and prescription focuses primarily on the attributes and activities of political representatives, whose responsibility it is to include, through their activities of rule, all ‘affected’ or ‘subjected’ populations with democratic claims to representation as members of the ‘demos’.Footnote ³³ The capacity of represented populations to exercise collective judgement and influence as democratic agents within representative processes has received much less attention: although the question of constituency agency has received some attention in broader debates on the constitution of the ‘demos’ in global democratic practice,Footnote ³⁴ its implications for the theory of global democratic representation have not been systematically examined.

An important question left open by much of this work concerns the capacity of representative practice to sustain democratic legitimacy in contemporary political governance institutions. Some recent work on international and transnational representation invokes the language of legitimacy – suggesting that some prescribed representative innovations, addressed to the remedy of democratic deficits, will also have the advantage of strengthening legitimacy.Footnote ³⁵ Such claims have merit insofar as representative practices can fulfil not only the inclusive political function of democraticity, but also the distinct political function of legitimization: resolving normative disputes over the institutional terms of political rule, through the subsumption of divergent individual judgements within a unified institutional framework for collective political agency.Footnote ³⁶ One key prerequisite for legitimizing functionality is that the parties to normative disputes over the terms of institutional rule – in democratic terms, those who claim a political ‘stake’ in its powers as affected parties,Footnote ³⁷ or those subjected to demands for institutional compliance or supportFootnote ³⁸ – are included in the collective democratic decision-making processes that offer some political remedy for these disputes. Insofar as representative innovations are successful in fostering such inclusion – through better aligning the boundaries of governance processes with the boundaries of collective democratic agency – their claim to be contributing to strengthened legitimacy in international institutions can be upheld.

The complex political function of legitimation requires more, however, than merely aligning the boundaries of represented constituencies with those of legitimacy problems – representing all democratic claimants in normative disputes over the institutional terms of governance power. For representative inclusion to contribute to remedying legitimacy problems, there is an additional requirement for represented populations to be actively engaged in representative processes, to the degree required for sustaining the legitimizing conversion of normative dispute among democratic claimants into collective democratic agency in the shared exercise of governing power (thus converting diffuse populations of democratic claimants into active democratic constituencies). With the shift from participatory to representative democratic models, the expectation for active political engagement of democratic constituents does not extend to their direct involvement in all the activities of political decision-making and implementation that constitute political rule itself. But to foster legitimacy through these democratic practices, representatives must still bring their constituents along through sufficient active political engagements to motivate and guide their support for governing institutions in the ways captured by the distinct normative concept of legitimacy.

More specifically, successful legitimation requires that represented groups have capacities to: recognize themselves as democratic constituents of ruling institutions, and as represented by participants in these governance processes;Footnote ³⁹ engage (communicatively or collaboratively) with one another, in creatively defining the common interests or problems that integrate them as members of democratic constituencies;Footnote ⁴⁰ and commit to political responsibilities for democratic burden-sharing and compromise, without which the legitimizing shift from normative contestation to democratic collaboration could not be motivationally sustained.Footnote ⁴¹ Agentic political capacities of these minimal kinds are acknowledged functional requirements already of some theoretical models of ‘formal’ political representation, which require represented constituents to engage collaboratively and responsibly in activities of authorization and accountability within their representative relationships.Footnote ⁴² As such, these formal representative instruments can make especially clear contributions to legitimation processes. But even with ‘substantive’ and ‘descriptive’ representative practices, which are more reliant on the agency of representatives to articulate and embody constituency interests and identities, legitimacy will still depend on constituents possessing these forms of active agency: without constituents’ active recognition, integration, and commitment operating in the background, representative rule would not convert to legitimizing support from institutions’ democratic claimants.

There is a longstanding and live democratic debate about whether representative legitimation of these kinds can realistically be achieved in political practice. An influential tradition of thought on democratic legitimacy – originating in Rousseau’s well-known claim that the democratic general will cannot be represented – contends that representative practices are incapable of empowering and sustaining adequate capabilities of these kinds for the exercise of legitimizing democratic agency.Footnote ⁴³ The contemporary political circumstances of transnational democratic projects produce some compelling grounds for supporting such sceptical views. Current configurations of global governance institutions – viewed systemically, as subsuming all governance levels and institutional types interacting within the overarching global order – are complex and dynamic, with uneven and shifting impacts on local and transnational populations.Footnote ⁴⁴ As a consequence, many populations of democratic claimants are themselves diffuse and dynamic – with contested and shifting boundaries, dispersed and diverse memberships, and weak political relationships to their purported representatives within governance institutions. This systemic disjuncture between populations of democratic claimants, and those connected socially and institutionally as active democratic constituencies, produces significant challenges for the development of legitimate representative instruments of global democracy.Footnote ⁴⁵

11.2.3 Constitutive Representation As Remedy for Transnational Legitimacy Problems

Given these challenges, proponents of representation as an instrument of global legitimation must consider further what instruments within representative practices can best foster and sustain the requisite dimensions of active democratic agency within represented constituencies. Here I propose that such practices can be helpfully conceptualized as a variant of what is sometimes called constitutive (or in more common terms, ‘constructivist’) representation.Footnote ⁴⁶ This idea captures the creative potential that arises in any representative process, which re-presents its subject in a new and different form from its original condition:Footnote ⁴⁷ not just describing or displacing a subject already adequately formed, but re-constituting the represented subject itself through the representative process. Established theoretical accounts of constitutive representation have been mostly concerned with the symbolic dimensions of representative practice – that is, the ways in which cultural categories of political interest and identity can be creatively constructed by representational activities of depiction, narration, and interpretation, prior to the wider ideational uptake of these categories by institutionally affected or subjected democratic claimants.Footnote ⁴⁸ Some theorists have deployed these representative concepts to capture the way self-appointed representatives can call forth or mobilize new constituencies in dynamic and contested governance contexts, such as those emerging in response to new transnational or global political problems.Footnote ⁴⁹

Here, I propose that this established concept of constitutive representation can be expanded to capture the practices through which representatives operate to constitute not only cultural categories of political membership, but further political capabilities for the exercise of legitimizing democratic agency. More specifically, we may consider how constitutive representation can function to cultivate, among members of fluid and emergent transnational constituencies, those capabilities for constituency self-recognition, communicative or collaborative engagement, and responsible political commitment, which we have established are required to support the active political agency that is the source of democratic legitimacy in representative practice. First, representatives may support the constitution of represented constituencies in the dimension of self-recognition by creatively interpreting and projecting latent or contested group identities – metaphorically, painting a portrait rather than holding up a mirror; in doing so, they may help constituents gain stronger recognition of their political status as a collective, and their democratic claims against international governance institutions. Second, representatives may support the constitution of represented constituencies in the dimension of communicative or collaborative engagement by supplying the interactive infrastructures and attentional salience required to bring disparate constituents together into creative political engagements of their own – building new and dynamic understandings of the content of shared interests in their governance institutions. And third, representatives may support the constitution of represented constituencies in the dimension of responsible political commitment by brokering engagements between their own represented constituents and other groups with competing represented interests – thus helping to contain partisanship among their represented subjects within the bounds and burdens of political responsibility to the wider democratic order.

More work needs to be done, however, to consider how such constitutive representation may function in practice in international institutions, to help remedy problems of legitimacy arising across the transnational constituencies of the global democratic order. It is accordingly to these more concrete political issues that we must turn our attention next; and in doing so, we will take up more directly the question of the role played by IOs in fostering these distinctive representative practices.

11.3.1 Orchestration as an Instrument of IO Representation

The traditional conceptual categories of modern political analysis are notoriously challenged by the theoretical task of capturing the distinctive forms of political agency embodied in IOs, as they now function within the complex governance systems of our contemporary global order. Modern theories of political power, legitimacy, and democracy have built conceptual apparatuses around the governance systems of sovereign States – drawing a clear conceptual line between the governing agency of political rule (centred on the lawmaking and enforcement powers of the State), and the democratic agency of the ruled (exercised collectively through representative or other practices of democratic legitimation). IOs do not fit neatly on either side of this conceptual divide, as they are both instruments and subjects of the sovereign powers of States, which invest them with their varied organizational powers. Moreover, any delegated ruling powers invested in IOs by States are subject further to the structural dynamics of a complex global governance system, in which a multiplicity of political agents must navigate and negotiate competing aims and powers to advance their governing or legitimizing aims.Footnote ⁵¹ Any theoretical account of IOs’ political agency within a democratic global order – whether focused on the side of governance or legitimation functions – must grapple with these systemic complexities.

These theoretical challenges have been the focus of significant attention from International Relations scholars seeking to account for the distinctive forms of political agency IOs exercise within processes of global governance – or in other terms, in activities of institutional rule. While many IOs still engage heavily in traditional State-centred forms of governance activity – such as negotiation of treaties, operation of interstate dispute resolution mechanisms, or monitoring of compliance – there is now wide recognition that their governance performance is often improved by utilization of soft and indirect forms of political agency, operating through the mobilization of ‘intermediary’ actors such as non-governmental organizations (NGOs), businesses, transgovernmental networks, transnational partnerships, and other IOs, in support of their governance goals. The concept of ‘orchestration’ is often used to describe these indirect modes of IO influence – serving as an umbrella concept for a wide variety of political strategies IOs use to support, facilitate, coordinate, and steer global governance activities via intermediary actors operating across a complex global institutional system.Footnote ⁵²

Some theoretical attention has been given to the question of how IO orchestration in global governance practices can itself be legitimized or democratized, as an instrument of institutional rule.Footnote ⁵³ But there has been less consideration of how the indirect forms of IO political agency captured by the orchestration concept may also help model IOs agency exercised on the other side of the democratic leger – in new kinds of representative practices that adapt democratic instruments to the complex social and institutional order of twenty-first century world politics. Here I propose that the concept of orchestration can instructively be applied in this latter way, to capture some important roles IOs can play in supporting, facilitating, coordinating, and steering a range of intermediary actors engaged in activities of constitutive representation – which cultivate in emergent constituencies, across the global order, the qualities required to support their roles in representative legitimation. Through so doing, IO orchestration activities can potentially support not only the functional efficacy of global governance practices, but also their democratic legitimacy.

11.3.2 How IOs Can Orchestrate the Constitution of Represented Constituencies

What, then, are some of the political activities through which IOs can orchestrate the constitution of represented constituencies, as democratic agents in international lawmaking processes? In exploring these activities, we can extrapolate and apply established theoretical work on the varied techniques of IO orchestration commonly deployed through intermediaries in support of global governance goals – exploring how these techniques may further be utilized in practices of constitutive representation.Footnote ⁵⁴ While IO orchestration capabilities and techniques are diverse, and vary greatly across contexts, existing orchestration scholarship has identified several with especially wide application: convening engagements among intermediaries, to catalyse collaborations; agenda-setting for programmes of collaborative work; financial and administrative assistance, to expand the material capabilities of intermediary actors; endorsement, to raise profiles or status of intermediary activities; and coordination, to increase collaborative synergies and impact.Footnote ⁵⁵ To map out how such orchestration techniques may be applicable within practices of constitutive representation, we can consider how they may operate to support the constitution of each of the legitimizing capabilities within transnational constituencies identified earlier: democratic self-recognition, engagement, and commitment.

First, IO orchestration activities may contribute to the constitution of democratic self-recognition within represented transnational constituencies – supporting the interpretation and projection of latent or contested identities among constituents who have been brought together not by longstanding historical trajectories of shared culture, but rather by more recent impositions of institutional affectedness or subjection producing a shared democratic claim. Two important orchestration techniques IOs can deploy to support these constitutive representative practices are convening and endorsement: here IOs can leverage their access to a range of (contextually varying, public or private) influential actors in their governance domains to identify representative intermediaries with inspirational visions and leadership capabilities; and they can further supply these representatives with the institutional platforms and networks required to project their visions across audiences wider institutional stakeholder constituencies. Administrative and financial assistance will typically be a further essential component of orchestration activities of this kind – whether this supports the convening of high profile representative public events or forums, or the brokerage of ongoing relationships between visionary representatives and their wider constituencies at more localized levels, which allow for wide political transmission of representative interpretations of collective democratic identities.

Second, IO orchestration activities may contribute to the constitution of communicative or collaborative engagement within represented transnational constituencies – supplying not only platforms for representatives to disseminate inspirational visions of the democratic identities and claims of their represented constituencies, but further opportunities for disparate members of transnational constituencies to contribute themselves to defining the political interests associated with these memberships. As above, orchestration techniques of convening and assistance can play important roles in supporting these constitutive practices. Here, however, the emphasis will be less on support for high profile platforming of representative visions, and more on support for deliberative network infrastructures and local community engagements diffused across the wider memberships of represented constituencies. To provide structure and focus for these diffused political processes of communicative and collaborative engagement, there may often be roles also for orchestration techniques of agenda-setting and coordination. Here, IOs can support the deliberative articulation of constituency interests by helping to frame questions, establish deliberative forums and processes, and create shared records of dialogues among members of transnational constituencies, in support of these legitimizing forms of collective democratic engagement.

Third, IO orchestration activities may contribute to the constitution of responsible political commitment within represented transnational constituencies – by fostering understandings not only of their internal democratic identities and interests, but also their political responsibilities to others across the wider global political order, with whom legitimate settlements must be reached through shared governance institutions. Supporting the constitution of wider political commitments of this kind may require an array of orchestration techniques – with agenda-setting playing an especially important role in framing representative engagements around problems and challenges of common global concern, as distinct from narrower partisan issues, and coordination helping to define clear norms for responsible political engagement of constituents with the wider governance institutions and communities through which their interests are to be advanced. These responsible political engagements can be aided further through building and disseminating more knowledge about constituencies’ mutual vulnerabilities and interdependencies, and through building capacity for secure and strong political leadership structures within represented constituencies. Here, assistance may also be a crucial feature of IO orchestration activity in this area – insofar as IOs’ financial, administrative, and technical resources can make valuable contributions to these knowledge- and capacity-building practices.

12.2.1 International Democratic Legitimacy

The representation of peoplesFootnote ²¹ by IOs is a question of political legitimacy. By political legitimacy, we mean a form of legitimacy that is not descriptive or subjective and cannot be reduced to popular acceptability, but is normative or objective.

More specifically, it is a form of normative procedural legitimacy we are concerned about. It focuses on the procedural aspects of lawmaking and should therefore be distinguished from more substantive forms of normative political legitimacy, such as justice. This brings us to another well-known distinction in the realm of political legitimacy: the distinction between input and output legitimacy. Theories of input legitimacy claim that the normative legitimacy of law derives from a combination of prior facts considered as ‘inputs’ into the lawmaking process. By contrast, output legitimacy refers to the substantive merits of international legal norms, be it their substantive justice or efficiency or both and therefore focuses on the ‘outputs’ of the lawmaking process. With its focus on procedural normative legitimacy,Footnote ²² this chapter sides, even if it objects to the terminology and to the distinction itself between output and input, with input accounts of political legitimacy.

This chapter understands political legitimacy, in quite a standard way, to refer to the normative questions of who has the right to rule (or the right to adopt legal norms or participate in the adoption of such norms) and how such a right to rule should be exerted in order to generate obligations for those subject to such rule.Footnote ²³ Under this standard account, these questions have usually been given a democratic response, and it is the case in this chapter as well. Only democratic political systems, in very broad terms and perhaps with a few exceptions, are regarded as legitimate, while authoritarian or mixed regimes are considered illegitimate. This means, still broadly speaking, that lawmaking should ultimately be able to be attributed to the people subject to the law,Footnote ²⁴ with some additional democratic requirements we will revert to below.

Input legitimacy in general, and democratic legitimacy in particular, as they are understood here, are regarded as the standard way of legitimizing domestic law and institutions. On the international plane, however, discussions have taken a somehow different turn. In the last few decades, indeed, many authors have assumed that democratic principles cannot apply to international lawmaking processes. This has brought them to endorse output legitimacy as an alternative way to provide a normative justification of international law.Footnote ²⁵

This exclusive focus on the output legitimacy of international law and institutions is problematic. There is no doubt that the substantive quality of international law is a matter of utmost importance, including for its normative justification. The main problem with output legitimacy, however, is that it identifies legitimacy entirely with either justice or efficiency. By doing so, it can no longer address the distinctive and important normative question of who should rule and how. When the sole focus of legitimacy is on output, indeed, the only way to address that question becomes instrumental. If the unique goal is to maximize the probability of adopting good (e.g. just or efficient) laws, the answer can only be that such laws should be adopted by those with the most capacity to do so. Important as this consideration might be in deciding who should have representative power to make public decisions in the first place, it does not resolve the problem of whom those representatives should be accountable to and hence of whom they should represent.

As a matter of fact, considerations of output legitimacy do not exhaust our concerns about the legitimacy of law and institutions domestically. Whatever should be the exact balance between expertocracy and democracy in any particular context, indeed, our concerns about normative legitimacy at the domestic level are never exhausted by considerations of output legitimacy. The reason is that our concerns about input legitimacy are considered as self-standing and not only instrumental to output. Given the identity of the ultimate individual subjects to the authority of both domestic and international law, that is, the same peoples albeit re-instituted in different publics, those considerations about the self-standing importance of input legitimacy should also apply to the political legitimacy of international law.Footnote ²⁶

In this chapter, we propose that the democratic legitimacy of any lawmaking institution should be assessed by reference to four basic, scalar abstract principles common to most accounts of democracy.Footnote ²⁷ First of all, the principle of ultimate, effective popular control. It derives directly from the ideal of self-government or popular sovereignty: all peoples subjected to law should have a relevant say in the process of making that law. They may, of course, confer that power to representatives. The peoples should, however, retain ultimate, effective control over their representatives (who may otherwise only claim to be such) and, through them, over institutions and decision-making processes in order to make self-government possible. Second, the principle of political equality. The people represented should have an equal say, directly or through their representatives, in holding that ultimate power of control. That means that no citizen or group of citizens should be able to impose their views unilaterally or have significantly greater political power to determine the law than others. Third, the principle of deliberative contestability. The peoples – or their representatives – should be able to contest, through deliberation, all the laws and decisions made. They should also have the capacity to engage in deliberative interaction with each other, thus promoting public (formal and informal) debate. Fourth, the principle of human rights’ protection. Individuals’ human rights that are constitutive of their basic moral equality and enable them to exercise ultimate control should also be protected in lawmaking processes and institutions in order for the latter to be democratically legitimate.

How do these four democratic principles apply to international institutions in general, and to IOs in particular?

The first democratic principle is quite clear in this respect, and it is the one which we will pay attention to in this chapter.Footnote ²⁸ Given that the peoples of the world cannot rule directly at the international level – in fact, they cannot do so at the domestic level either – they should confer their power to rule to representative institutions and to those institutions’ organs. This turns the question of the democratic legitimacy of international law into one of democratic representation. Accordingly, the assessment of the democratic legitimacy of institutions of international lawmaking crucially depends on whether those institutions, and IOs in particular, can be said to have been conferred their powers by the peoples subjected to the decisions to be made (the so-called ‘authorization’) and, even more importantly, on whether the peoples may access effective mechanisms of ultimate control over their representatives and over the decision-making system as a whole (the so-called ‘control’).

It is important to emphasize that representation would be required even if direct popular participation were feasible internationally: not only does democratic participation require the primary institution of the people as such and hence a form of representation of that people in the first place, but representation is also required to guarantee political equality and to organize equal participation and deliberation all along. Contrary to what one often reads, therefore, the relationship between the concepts of democracy and representation is not instrumental and contingent, but intrinsic and constitutive.Footnote ²⁹ This is an important point because various forms of compensatory participation and deliberation are often put forward precisely in lieu of international representation that is held too difficult to organize by many authorsFootnote ³⁰ – usually, moreover, because it is also wrongly identified with and reduced to electoral representation. What those authors do not understand, however, is that representation and participation are in a mutual and constant relationship to the extent that the represented should also be able to participate in representative lawmaking processes and deliberate continuously with their representatives and vice-versa for the latter to represent them.Footnote ³¹

In our traditional (modern) understanding of representative democracy, at least when applied to domestic lawmaking, these four principles have required the existence of a parliament with fundamental lawmaking powers whose members are elected democratically, a directly or indirectly elected executive and an independent and self-standing judiciary, among other democratic institutions. Nothing like this, of course, exists (yet) at the international level, including in most IOs. This is even the case in the European Union (EU) where the directly elected parliamentary organ of the EU holds neither the lawmaking initiative nor the legislative monopoly and does not elect either of the two executive organs of the EU. To that extent, the proposed conception of the democratic legitimacy of IOs cannot be that of a worldwide or global democratic State with a single global people or instituted public represented by various directly or indirectly elected officials organized in parliamentary, executive, and judicial authorities.Footnote ³² As a result, whether or not one considers that such a cosmopolitan and monist account of international democratic representation is justifiable under ideal circumstances, it is clearly far from being possible in present, deeply pluralist, real conditions. Importantly, however, this does not imply that the principles of democratic legitimacy cannot be respected through other institutional configurations. Those principles albeit developed for States are actually general enough to adapt to the international institutional context.

Before turning, in the next sections, to the crucial two questions of whom IOs should ‘democratically’ represent and how, it is important to briefly specify what democratic ‘representation’ amounts to more specifically.

12.2.2 International Democratic Representation

The conceptions of political and especially democratic representation we use today are very recent: they are modern, even if earlier traces may be found, of course.Footnote ³³ The origin of these modern conceptions, however, is medieval. They arose from numerous theological and legal permutations that took place in Europe from the twelfth century onwards.

These so-called ‘political-legal’ conceptions of representation originated in the reception of Roman private law and constitute the core of the medieval jus commune of representation.Footnote ³⁴ In short, they developed on the basis of two conceptions of representation elaborated by lawyers, first in canon law and then in civil law, especially in the Germanic and Italian traditions thereof.

The first conception of representation relies on the notion of corporation or legal person such as a city or the Church (and the pars pro toto representation by this incorporated or legal person and, by extension, the representation of this legal person by a particular physical person therein, that is, its ‘head’ – caput – or ‘organ’). The most famous discussion thereof was that of Marsilius of Padua in the thirteenth century.Footnote ³⁵ We speak here of ‘personal representation’ or ‘incorporation-based representation’ because the representative acts ‘as’ or ‘like’ the represented: it stands as them (stand as representation). The second medieval conception of representation is that of the mandate or legal delegation (and the principal-agent relation) developed in particular by Bartolus of Saxoferrato in the fourteenth century.Footnote ³⁶ We speak here of ‘conventional representation’ or ‘agency-based representation’ because the representative acts ‘for’ the represented: it acts in their place (act for representation).

With the advent of modernity, this medieval private law of representation gave rise, in domestic law, to the public law of political representation. The latter still bears the traces of the former’s two conceptions of representation, however.Footnote ³⁷ The first conception of incorporation-based representation gave rise, in the seventeenth century, to the political notion of the person of the State and its sovereign, and then of the sovereign State itself and its organs. The second conception, that of agency-based or delegation-based representation, gave rise to the democratic notion of the electoral and, more generally, to the political ‘mandate’ of the representative from the nineteenth century onwards. As we will explain in Section 12.4, the international law of representation actually bears the traces of the medieval private law of representation even more directly than domestic law.

A common feature of the various contemporary conceptions of democratic representation that ensued from this modern understanding is the idea that representatives should take into account and promote the interests of the represented. This is clearly the case of contemporary agency-based conceptions in which the agent should act not only in the name of the represented (‘as’ them), but also to their benefit (‘for’ them).Footnote ³⁸ However, it is also true of the so-called ‘symbolic’ and ‘descriptive’ conceptions of representation according to which the representative relationship depends on other, constitutive features that are related to the protection of their interests.Footnote ³⁹ It even applies to the so-called ‘constructive’ or ‘claim’ conceptions of representation, according to which the only definitional requirement of representation is that an agent claims to represent another one in terms that might prove acceptable through a constructive process of the represented’s interests and this, whether the latter have authorized and controlled those terms or not.Footnote ⁴⁰

To revert to our earlier discussion of input and output legitimacy, the fact that representatives should take into account and promote the interests of the represented in most contemporary accounts of democratic representationFootnote ⁴¹ does not, however, imply that the corresponding notion of political legitimacy is necessarily a substantive one and based on output only. Nor does it imply that input legitimacy should only be understood in instrumental terms. Many modern views of political representation, starting with Thomas Hobbes’,Footnote ⁴² have indeed made the latter dependent on formal acts of authorization, whether ex ante or ex post. This is the case, for instance, of the so-called ‘trustee’ and ‘delegate’ models of representation.Footnote ⁴³ In both cases, representatives act for the represented on the basis of some kind of authorization. They differ, however, in the degree of autonomy that representatives should enjoy in order to effectively represent the represented. While the delegate model claims that representatives should operate under the tight terms of an explicit act of delegation or mandate, the trustee model relies on the representative’s relative independence in order to best protect the interests of the represented.

Accordingly, and regardless of how those debates have evolved domestically and without entering into those details here, there are two conditions which political representation should comply with in order to be regarded as democratically legitimate, and which are of utmost importance to the institutional organization of international representation as well.

The first one derives from the principle of ultimate, effective popular control. Representatives should be authorized by the people they are meant to represent, and should also be ultimately and effectively controlled by those people.Footnote ⁴⁴ Independently of the degree of autonomy representatives might be granted, the represented should always possess effective mechanisms to hold the representatives accountable and retain ultimate control over their actions. Periodic elections with electoral mandate have proven to be a relatively successful institutional mechanism to do so. However, they are neither a necessary nor a sufficient condition of democratic legitimacy. Indeed, elections do not necessarily guarantee ultimate, effective control by the people in domestic democratic systems. Moreover, there might be other mechanisms that could complement or even replace elections as instruments of ultimate, effective popular control. Many modern conceptions of representation have been excessively dependent on electoral accountability by making it a conceptual requirement of democratic legitimacy.Footnote ⁴⁵ As a result, in contexts where such electoral control does not yet or simply cannot exist or is difficult to organize, as it is on the international plane, this has brought many to give up on democratic representation or, by extension, on democratic legitimacy altogether, as mentioned before.

The second condition is a consequence of the first one and is even more crucial to international representation than it is domestically. Democratic representation inherently possesses an instituting dimension. This has often been overlooked at the domestic level because the institutional framework of the State is long-lasting and usually taken for granted when organizing democratic representation. Political representation can only occur through law and institutions. It follows that democratic representatives should operate in institutional and legal frameworks and should be instituted themselves as such by those legal frameworks. To revert to a point made before, the way in which representatives are instituted and hence constructed is actually a mutual constructive process that also includes the institutional construction of the represented themselves as a people or a public in return.Footnote ⁴⁶ Importantly, therefore, and even though there is something that has long been identified as paradoxical about it,Footnote ⁴⁷ that process should also fall under the principle of ultimate, effective control and be adequately institutionalized. This is actually why the instituting dimension of democratic representation should not be separated from the authorizing one, and vice-versa.Footnote ⁴⁸ This is another reason why participation and representation should always go hand in hand, as argued before.

To recapitulate, then, this chapter approaches democratic representation as a relation that institutes, authorizes, and holds power accountable under the ultimate control of the people and thereby contributes to the justification of political authority, including of lawmaking authority. It gives the representative a claim to act as and for the represented, leading to the attribution of those acts to the represented if the conditions of authorization and control are fulfilled.

12.4.1 Against Functional Representation by International Organizations: Input Matters

To date, representation in the international law of IOs has been approached mostly as a form of instrumental representation of States’ functions and framed as providing IOs with output legitimacy. To that extent, the legitimacy of IOs and the representation by IOs may be described as ‘functional’.

Things are more complicated than that, however, and there is a certain ambivalence nowadays between functional and democratic legitimacy in IO law and hence between output and input legitimacy. One may distinguish two periods in this respect: a first one that goes from the foundation of the early IOs in the nineteenth century, and a second one that opens with the end of the Cold War.Footnote ⁶² While the second period has brought in a new emphasis on democratic input legitimacy, it has not succeeded in trumping functional output legitimacy entirely. As a result, both concerns coexist today in the international law of representation by IOs.

First period: the carelessness of functional good governance. When the first IOs were established in the late nineteenth century, they were to remain under the sovereign control of their Member States. Conceived from a functionalist perspective, IOs were simply to perform the ‘functions’ that their Member States had ‘delegated’ to them, as opposed to exercising the ‘sovereign powers’ those States could have ‘conferred’ to them.

In this context, expecting IOs to conform to the principles of democratic good government and represent their Member States’ peoples would have been contradictory. Indeed, to the extent that States were to remain in charge of the sovereign ‘government’ of their peoples, IOs were not designed to ‘govern’ those peoples. According to that conception, if IOs were to be described as ‘public’, it was only so in a functional or instrumental way.Footnote ⁶³ They could be considered public only insofar as they had been established to serve the ‘public functions’ of their Member States. As for their alleged ‘public authority’, it amounted at most to the exercise of a de facto authority. There was no claim that that authority was exercised on behalf of an instituted public that would have authorized its exercise, nor was it to be controlled by that public in return, a public that could not therefore be considered as the ‘author’ of IO law.

Clearly, then, and despite many rhetorical preambles (such as the United Nations (UN) Charter’s reference to ‘We the peoples of the United Nations’), IOs have not yet been organized or ordered by their Member States’ peoples as political institutions representing them. Worse, many IOs have actually been organized by their States to remain ‘apolitical’.Footnote ⁶⁴ This explains why IO Member States did not initially consider it necessary to insert rules and principles of public law into the internal law of their IOs in order to organize the latter’s good government, including their democratic representativeness.Footnote ⁶⁵ For example, there was and still is no trace of a principle of separation of powers and even less of a democratic principle in most IOs but the EU.

No wonder then that the representation by IOs has mostly been organized so as to follow a functional approach to legitimacy by output: it should ensure the implementation of State functions and hence the protection of their Member States’ interests. When it is exceptionally framed as input legitimacy, which is also sometimes the case, representation by IOs remains functional, however. It relies on the IO’s organs’ expertise, for instance, in implementing its Member States’ functions and protecting their interests. The same applies to the representation of States or even of CSOs in IOs. The latter does not necessarily seek to represent the peoples of those States or the civil society organized through those CSOs. Their representativeness is measured by reference to certain characteristics of States or CSOs that may be of an instrumental purpose for the IO, such as the economic, military, or financial power of some of its Member States (usually a few of them)Footnote ⁶⁶ and the expertise and financial resources of certain CSOs.Footnote ⁶⁷

This is problematic from a democratic legitimacy perspective that places emphasis on the distinct relevance of input legitimacy, as argued in Section 12.2.1. Even more so as the protection of the represented’s interests through representation does not exclude the relevance of input legitimacy through authorization and control, as argued in Section 12.2.2. Important as it is to take into account the quality of output under such instrumental approaches to representation by IOs and to rely on experts within those IOs’ organs when making law, securing ultimate, effective popular control over our representatives should also matter from a non-instrumental perspective of democratic legitimacy. Following Ernst-Wolfgang Böckenförde, one should consider democratic representation as a form of ‘material’ rather than merely ‘formal’ representation of pre-existing interests.Footnote ⁶⁸ Representation enables the people(s) to specify, through their institutions’ deliberation, what actually amounts to their public interest. One may argue therefore that the international ‘public’ and, by extension, its ‘public interest’ does not actually pre-exist international political representation.

Importantly, there was a turning point in the 1990s and a second period opened in the international law of IOs: the call for democratic good government. It was during that period indeed that many IOs became more autonomous in their international lawmaking procedures, subjecting States and, directly or indirectly, their peoples to their legal authority. Hence the growing need to justify that authority and hence to control their power in order to make it legitimate, including through democratic representation.

The end of the Cold War also corresponded to the rekindling of international democracy law, including in its application to the international institutional order.Footnote ⁶⁹ The specific obligation of IO Member States to organize their IOs in a democratic manner may be grounded in the human right to democratic participation under Article 25(a) of the International Covenant on Civil and Political Rights.Footnote ⁷⁰ Unfortunately, States are still deeply divided over the scope of application of those international democracy law obligations: Southern and Eastern States oppose its domestic application inside States and require more international democracy including in IOs, while Northern and Western States oppose its international application including inside IOs.Footnote ⁷¹

Unsurprisingly, therefore, it is the new avatar of good government in domestic public law, that is, the public-private approach to ‘good governance’, that has been transposed to IOs.Footnote ⁷² This explains how the so-called process of ‘constitutionalization’ of IOsFootnote ⁷³ has in fact largely turned into an exercise of technocratic ‘rationalization’.Footnote ⁷⁴

What makes this current system of ‘techno-democratic’Footnote ⁷⁵ governance of IOs difficult to criticize, however, is that the appearances of democratic legitimacy have been saved. Terms are regularly borrowed from the lexicon of input legitimacy, even if they are deprived of their original meaning. This is what Ayelet Berman has referred to as the ‘false sense of legitimacy’ of IOs.Footnote ⁷⁶

To be sure, ‘participation’ or ‘inclusion’ in IOs together with their ‘accountability’ are often emphasized in those proposals. However, and to start with the term ‘participation’, it is mostly used to refer to various types of partnerships with ‘non-State actors’ including those from the private sector, and not to the participation of citizens themselves.Footnote ⁷⁷ It does not, moreover, and curiously, concern itself with the representativeness of those so-called ‘participants’. Those non-State actors such as CSOs cannot indeed claim to participate in and of themselves and should somehow be related to the segment of the civil society they are contributing to organize and hence to represent. More generally, as mentioned earlier in the chapter, there can be no democratic participation without some form of representation in order to guarantee political equality and organize self-government.

As to the term ‘inclusion’, second, it usually refers to ‘stakeholders’ in that context, that is, pre-established interest groups.Footnote ⁷⁸ Unlike the latter social groups, however, peoples institute themselves through their representatives in order to participate in their self-government and thereby to emancipate themselves as publics from their prior collectives.Footnote ⁷⁹ And the same may be said of CSOs (e.g. trade unions) whose representative self-organization contributes to transform prior social groups (e.g. workers). Finally, ‘accountability’ is a curious in-between.Footnote ⁸⁰ Its procedures are not usually approached as political unlike authorization or election, on the one hand, and its consequences are not political or even legal unlike responsibility, on the other.Footnote ⁸¹ It is achieved through various ‘audits’ and other self-referential effectiveness assessment mechanisms stemming from business management that allow IOs to provide evidence of their ‘compliance’ with ‘good governance’ standards.Footnote ⁸²

Ironically for our purpose in this chapter, ‘representation’ is not usually mentioned expressly in this ‘window dressing’ exercise in the legitimation of IOs. This may be because of the ambivalence between functional output legitimacy and democratic input legitimacy: democratic representation requires popular input and hence ultimate, effective popular control of the represented over the representatives.

Clarifying this is particularly pressing as alternative forms of international democratic representation are currently discussed in democratic theory that are more output oriented and instrumental to the representation of interests. It is the case, for instance, of ‘mimetic’ or, at least, ‘statistic’ representation,Footnote ⁸³ ‘altruistic’ representationFootnote ⁸⁴ and/or ‘epistemic’ representation.Footnote ⁸⁵ All three purport to represent pre-identified traits or interests to be promoted (functional output) or, at least, knowledge or expertise in how to do so (functional input). Those forms of representation are gaining traction in discussions of international democracy as well.Footnote ⁸⁶ Not only, indeed, do they fit the prevalent functionalist approach to representation by IOs and its output legitimacy framework just described. They also claim to vest it with a democratic justification without, however, reverting to the proposed conception of democratic representation that requires authorization and control. This may be because the proponents of those alternative forms of democratic representation wrongly conflate the latter conception with electoral representation. In light of the many hurdles in the organization of international elections, they then disparage ultimate, effective popular control from the realm of realistic possibilities in international representation and hence revert to alternatives.

Whatever the virtues of those alternative forms of representation as complements to international democratic representation, they cannot be regarded as democratically legitimate in themselves, and actually amount to illegitimate ‘shortcuts’, to cite Cristina Lafont.Footnote ⁸⁷ Those alternative forms of representation should never be regarded as substitutes for wider democratic deliberation and hence representation in the strong democratic sense. They cannot be regarded as legitimate in the absence of explicit and effective forms of authorization and control. Indeed, what authorization rules out, first, is the self-appointment of the representative, including in the form of self-representation by mimetism or statistical reproduction. The requirement of authorization also excludes the random selection of representatives by a third party (or by lot) unless there has been some form of ex ante authorization to that random selection. Second, control implies that the represented disposes of ultimate, effective means of supervision over the activity of the representative. The mere ‘claim’ to represent is not enough to respect the equal autonomy of the represented citizens, therefore.Footnote ⁸⁸

12.4.2 Against Incorporation and Agency Representation by International Organizations: Publicness Matters

If the international law of representation has generally made very little room for the political dimension of international representation, it is also because the private law analogies that have prevailed in the moulding of the modern law of public representation domestically have remained even more prevalent in international law. As a matter of fact, the public turn has not yet taken place in the international law of representation.

The reason for this is very simple: from the twelfth century onwards, and since the first diplomatic exchanges between Italian cities, the international law of representation has developed as a private law of representation in the mould of the nascent medieval jus commune of representation.Footnote ⁸⁹ This was confirmed by the consolidation of classical international law in the seventeenth century, an international law conceived, yet again and more generally, on a private law model and which approached the rights and obligations of States by analogy with those of private individuals.

This actually echoes a more general feature of the international legal order. Because of what one of us has referred to as international law’s ‘methodological statism’, the international law of statehood has eluded the political origin of the State and its publicity is simply built upon domestic public law.Footnote ⁹⁰ As a result, until this day and most of the time, the public and, by extension, the political dimension of States has been ignored in the legal organization of their international representation. It has been relegated to the realm of domestic public law and hardly any corresponding international public law has developed on the issue. Sometimes, the question of the representativeness of States’ international representatives is even relegated to the side, as a matter of ‘fact’ rather than law.Footnote ⁹¹ States being the ‘original persons’ of international law, it is their own representation that is at stake qua persons as opposed to public institutions, by analogy with legal persons of private law.

In turn, this explains why the international law of representation has been more concerned with the representatives of States than by States themselves as representatives – or only as representatives of other States. By analogy, this is also what applies to the representation of IOs rather than by IOs in the international law of representation. IOs are represented as ‘persons’ under international law as opposed to public institutions. Their representatives are not considered to represent peoples, as a result, hence actually the equivocation between IO ‘organs’ and their ‘agents’.

To be sure, the question of representation by IOs also arises in international law. It does so only to address the representation of States by IOs, however, and not that of those States’ peoples by those IOs. Even when IOs represent their Member States, indeed, and with exceptions such as the EU, IO law does not also position those IOs as representatives of the peoples of these States. Sometimes, of course, the representation of these States in their IOs’ interstate organs requires that State representatives should also represent their peoples democratically. It is then organized by the internal law of some IOs so as to be democratic. One may think here of certain IO’s parliamentary interstate bodies or of the democratic credentials required of State representatives of certain IOs at the time of accreditation or even for purposes of State accession/suspension/exclusion of the IOs.Footnote ⁹²

In all these cases, however, it is the representation by States in those IOs that should be democratic, that is, what we have referred earlier to as democratic representation in and through IOs, and not the representation by the IOs themselves – even if the former contributes to the latter, of course.

Again, the reason for this is straightforward under international law. An IO is considered a legal person distinct from its Member States and established precisely to represent its Member States. IOs are not usually ordered so as to institute the peoples of their Member States as new public institutions and, as such, are not considered to represent those peoples. This is considered to be the price of safeguarding their Member States’ sovereignty and, by extension, sometimes those States’ peoples’ popular sovereignty, albeit wrongly so as we argued earlier.Footnote ⁹³

Just as the international law of representation of States by States, the international law of representation of States by IOs has been articulated around two models originating from the medieval jus commune of representation: conventional and personal representation. In what follows, we will assess them in turn, before criticizing their lack of combination in a third step.

First, conventional or agency-based representation. This type of international representation is inspired by the private law of agency and articulates the relation of delegation between a principal and an agent.

Conventional representation applies to representation between any persons or subjects of international law and, by extension, between IOs.Footnote ⁹⁴ Conventional representation sometimes also applies to the representation of States by IOs. This is only the case, however, insofar as it does not concern relations of ‘membership’ of those States in the IO. Those relations, indeed, are in principle covered by the internal law of the IO, to the extent that the activity of the IO falls within the scope of the powers that its Member States have conferred to it. In such cases, it is the second type of personal or incorporation-based representation of States by the IO that applies. We will revert to this alleged incompatibility and tension between the two approaches to representation at the end of this section.

Because conventional representation works along the lines of a delegation or mandate between two private persons, it does not take into consideration the representative dimension of these ‘persons’ themselves and hence their institutional dimension. And this, even if the represented person is also a public institution under domestic law and a democratic State as it is or should be the case of IO Member States. This is highly regrettable from a democratic point of view.

As a matter of fact, resorting to principal-agent delegation as a model for democratic representation has been criticized in a domestic context as well.Footnote ⁹⁵ Indeed, despite its historical origins, many authors consider that the relationship of political representation should now be distinguished from a relationship of mandate, delegation, or agency in private law. Among the difficulties raised by the agency analogy, one should mention, first of all, its reliance on the pre-existing legal personalities of both the representatives and the represented and, further, on their independence, as opposed to their mutual construction as political institutions, that is, both as representative institutions and as their peoples. A second difficulty pertains to the agency analogy’s reliance on a one-way relationship of delegation from the principal to the agent and its absence of mutuality, as opposed to the constant participation and to-ing and fro-ing between political representatives and represented. Finally, a third difficulty lies in the agency model’s voluntarist approach to authorization by simple consent or veto, as opposed to other modes of political authorization and control, for instance through participation.Footnote ⁹⁶

Once applied to IOs, those three difficulties are magnified. First of all, by treating IOs and States as distinct and independent persons of international law as opposed to composite institutions, the conventional model of representation neglects not only the institutional continuity between IOs and their Member States, but also their institutional ties to the publics they institute and re-institute. In fact, applying a private law analogy to the international representation by IOs is even more problematic than it is in the case of the international representation by States. There is indeed no internal IO public law to organize the representation by those IOs and hence no background legal and institutional framework. Merely resorting to analogies with Member States’ domestic public law as a complement only makes things worse in this respect.Footnote ⁹⁷ Indeed, those public law analogies suffer from the same neglect of the composite institutional nature of IOs for they treat the latter as public institutions separate from their Member States and place them on an equal plane as comparable institutions.Footnote ⁹⁸ Second, and it is related, the agency-based approach to representation by IOs, by conflating public and private powers, cannot prevent the confusion between the conferral of powers by States to IOs and their transfer by IOs to private persons. As a matter of fact, the confusion has been enhanced by the generalized recourse to the private law term ‘delegation’ to cover both cases,Footnote ⁹⁹ and the public-private interchangeability of ‘functions’ the term seems to condone. The privatization of IO powers today may therefore be attributed in great part to the conventional understanding of representation by IOs.

The second type of international representation of States by IOs one encounters in the international law of representation is personal or incorporation-based representation. It relies on the second private law dimension of the medieval jus commune of representation, that is, the law of legal persons or corporations.

It is the most common of the two models of representation by IOs. Indeed, the IO and its distinct personality are not organized separately from its Member States for their own sake, but in order to personify the ensemble of its Member States and hence to represent them so as to bind them together through the internal and external actions of its organs. Still, one should not go as far as to resort here to the term ‘institutional’ representation used by Évelyne LagrangeFootnote ¹⁰⁰ when referring to this understanding of the representation by IOs. And this, in order precisely to avoid any confusion with the representation that institutes or rather re-institutes a given IO’s Member States’ peoples as new publics. With some exceptions, such as the EU, indeed, the international law of IOs does not yet consider States’ peoples as publics to re-institute and then represent in international law in the manner of new publics and sovereign international institutions. It does not treat IO Member States as re-instituted public institutions,Footnote ¹⁰¹ but only by analogy to private persons who become members of a private corporation and are re-personified as such.

This missing public institutional dimension of representation by IOs is the key objection to the personal model. It undermines their democratic legitimacy, especially with respect to ultimate, effective popular control. The incorporation-based representation of States by their IOs throws a ‘democratic veil’ over States at the same time as it knits a corporate or personal veil around IOs.Footnote ¹⁰² That veil interrupts the relationship of democratic representation that should link an IO’s Member States to their peoples by interposing the screen of the IO’s legal personality.

Besides those two critiques against the conventional and personal approaches to the representation by IOs separately, the most important and third set of critiques actually pertains to their lack of combination.

As mentioned in Section 12.2.2, democratic representation when it institutes and then authorizes and controls political power is a relation of mediation: it does not only bring together the represented and the representatives, but it also separates them at the same time.Footnote ¹⁰³ When it institutes this power, democratic representation therefore constructs, by separating them, the subjects of representation which do not pre-exist their representation. It is through the institution of the State, indeed, that the State’s people is instituted into a public. When, in a second step, democratic representation enables the represented to authorize and control the exercise of instituted power, it makes it possible to separate, once again, the representatives (this time, the parliamentarians or the government) from the represented (this time, the citizens or the governed) while tying them to each other.

Democratic representation thereby brings together two subjects in two successive and intertwined relations: first, it links the State, or any other institution of public law, to the people which the State institutes as a public, contributes to create as such and consequently represents (this is what one may refer to as ‘institution-representation’, whose ancestor is incorporation-based representation or ‘stand as’ representation); and, then, in a second stage, democratic representation connects the representatives of that State, that is to say its organs, to the represented people, that is to say the citizens (this is what one may refer to as ‘authorization-representation’, whose ancestor is the agency-based representation-mandate or ‘act for’ representation). Both dimensions are essential to democratic representation. Too often, indeed, contemporary discussions of domestic democratic representation are torn between two extremes: either they reduce representation, following a constructivist and almost neo-Hobbesian tendency, to the moment of the institution of the people without regard to the need for constant re-authorization and control thereafter between representatives and represented as guarantors of self-government and political equality;Footnote ¹⁰⁴ or they reduce representation to the electoral mandate, following an ontological and neo-Schmittian move that ignores all too quickly that the people and the citizens do not pre-exist the figuration of representation and must first be instituted as such.Footnote ¹⁰⁵

In fact, it is the characteristic of the State’s internal constitutional framework, a framework that is destined to last, that explains how easily it has become today to forget the first instituting dimension of representation in favour of the second, that is, the representation-authorization, which takes place much more regularly. Providentially, international law and its complex institutional framework gives us the opportunity to capture them together again. The multiplicity of representative institutions in the international legal order helps us re-connect the two dimensions of democratic representation and make them both visible again. Thereby, international representation by IOs may contribute to counter what some critiques have referred to as the ‘flattening’ of the tridimensional relation of political representation and its reduction to a bidimensional and horizontal authorizing dimension.Footnote ¹⁰⁶

For this to be the case, however, the instituting dimension of the personal model of representation by IOs and the authorizing dimension of the conventional model need not only to be fixed in themselves, as we proposed earlier in this section, but also brought together in future interpretations of the international law of representation by IOs. At the moment, indeed, they coexist as juxtaposed approaches to representation by IOs in the international law of representation without being articulated with one another.Footnote ¹⁰⁷ The instituted IO is exempted from the need for further authorization and control, thereby giving rise to a risk of domination; and the authorized IO is taken to exist without a prior institution, thereby giving rise to a risk of political inequality. Worse, both dimensions of democratic representation are also opposed at times, and treated as incompatible as mentioned earlier in this section.

13.2.1 The Unity of the Represented

The first difficulty when thinking about democratic representation in the EU is to identify the entity represented. Most political theories of representation view it as an operation whereby a more or less complex authority expresses a position on behalf of another entity, generally considered both unique and ultimate. For instance, Michel Troper and Manuela Albertone try to identify a ‘meta-concept’ of political representation with the following two elements: ‘on the one hand, the capacity of an authority to want or act on behalf of another entity to which that will and those acts are imputed. On the other hand, the supremacy of that authority and the function it exercises.’Footnote ⁹ The characteristics of the entity represented – its unity and superiority – are conceived around the idea that the operation of representation would precisely aim at making an identifiable totality express itself, or even exist. According to these authors, the legitimizing function of representation in political theory consists ‘in justifying power by claiming that it is exercised in another’s name’,Footnote ¹⁰ so that rulers would not act in their own name, but on behalf of another higher entity, namely – in democratic theory – the demos. The unitary character of the represented is particularly problematic because it is based on a fiction, or, at the very least, a stripped-down reality. Whereas, in the monarchical era, the unity of the representative made it possible to embody the multitude and conceive it as a whole, the advent of the democratic ideal now postulates the unity of the represented.Footnote ¹¹ It is the case of the invention of the ‘nation’, which is generally conceived as a purely abstract entity, or even of a ‘people’, whose representation cannot account for the complexity of the society that composes it. Henceforth, the unity of the represented has become a logical requirement of political representation. Representation requires assuming the unity of the entity represented, and, therefore, ignoring the diversity and complexity of its composition and the conflictual relationships that run through it. If the represented was fragmented into several independent entities, it would be impossible to make it express itself or even exist as such. Representation presupposes the autonomy of the represented, whether it is a people, a nation, a society, regardless of its components.

13.2.2 The Issue of Federative Duality

This conceptual approach to representation that presupposes the unity of the represented raises an important issue in federative type groups which, like the EU, are defined by the duality and equality of the federal and federated levels. The combination of federated and federal wills can hardly be appended to that of an encompassing totality that would subsume them, without risking at the same time to dilute them in the new entity. This is why the thinkers of federative theory believe that ‘in a federal structure, there are, legally speaking, two peoples: the federated people and the federal people’.Footnote ¹² The notion of ‘people’, like that of sovereignty, becomes problematic when thinking about the duality inherent in federalism.

It is from this perspective that one must understand the ambiguity of Article 10 of the Treaty on European Union (TEU) on the democratic principles of the EU. Article 10(1) TEU states that ‘the functioning of the Union shall be founded on representative democracy’ and yet, no demos is identified in whose name the cratos would be exercised. In order to circumvent the difficulty of having to choose between a single people or multiple peoples as represented entities, Article 10(2) TEU designates two collective entities: first, the ‘citizens’, who are ‘directly represented at Union level in the European Parliament’; and second, the ‘Member States’, that are represented in the European Council and in the Council by representatives who are said to be ‘themselves democratically accountable either to their national Parliaments, or to their citizens’. The latter results from the Treaty of Lisbon, which, by taking the Councils out of the orbit of pure interstate representation and into the democratic field, breaks with the conception that tended to present the European Parliament as holding the monopoly of democratic representation within the institutional system of the Union.Footnote ¹³ Article 10(2) TEU recognizes the democratic nature of the representation of Member States in intergovernmental institutions by highlighting that it is the representation of the national political communities of the Member States that is ensured by that representation. This convoluted construction tries to restore the idea of the complex nature of the EU, which is sometimes referred to as a ‘democratic federation of States and citizens’ to express the duality it embodies.Footnote ¹⁴ The distinction between ‘direct’ (parliamentary) and ‘indirect’ (intergovernmental) forms of representation may seem surprising at first, since, by hypothesis, representation presupposes intermediation and is generally opposed to so-called ‘direct’ democracy (which itself is never really such). Only the need to emphasize the unsurpassable duality of the federative structure explains the need to resort to this distinction between ‘direct’ and ‘indirect’ democratic representation.

The dual approach to democratic representation, however, often gives rise to the impression that its direct and indirect forms are in competition. Democratic representation in the EU would operate according to a ‘connected vessels’ system,Footnote ¹⁵ in which the strengthening of direct democratic representation in Parliament would necessarily detract from indirect democratic representation in the Councils. For example, according to Dieter Grimm, ‘the transformation of the Union into a parliamentary system would weaken rather than strengthen democracy in Europe’, because it increases the risk ‘that a Member State will be subject to a law that has not been approved by its democratically elected and controlled organs’.Footnote ¹⁶ For this author, in the absence of a genuine European demos whose members imagine themselves as such, for want in particular of a European public space linking European citizens and their representative organs, any increase in the powers of the European Parliament, which can only lead to a corresponding decrease in the powers of the Councils, would weaken the democratic legitimacy of the Union, which essentially emanates from the national demoi and their control over the positions of their national representatives within the Councils. In other words, the strengthening of ‘direct’ democratic representation would necessarily be at the expense of ‘indirect’ democratic representation. Yet, in the event of a conflict, ‘indirect’ democratic representation, which is rooted in Member States, would be the true source of democratic legitimacy.Footnote ¹⁷

This is why the hybrid form of dual democratic representation of the federated and federal levels is perceived as conflictual, unstable, and unsatisfactory. The concept of democratic representation seems to be captive of the State and unitary context in which it was mainly deployed. How do we escape this aporia of the unity of the represented? One solution is to duplicate, rather than oppose, the two components of democratic representation in the European federative context, that is, the States and the citizens.

13.2.3 A Solution? Dual Political Citizenship

The temptation to oppose the two levels of democratic representation, or to play the federated people against the federal people, underestimates the ubiquity of the represented components. Each of them includes two sides: one federated, and the other federal. Because they each are twofold, they cannot fully oppose each other without contradicting themselves. Yet, the question of the duplication of each of the components of democratic representation does not arise in the same terms.

The duplication of the State component of representation is classic. EU States may act as ‘members’ of the Union, thus participating in its representation, but also as ‘monadic’ States, representing themselves. It therefore seems simplistic, even erroneous, to oppose the European and national levels of democratic representation in the EU, since the former already comes in part from the latter. State duplication means that, on the one hand, the Union can only be democratic if its Member States are democratic, and, on the other hand, the latter can only remain democratic if the Union is democratic too. Admittedly, the boundary between the two ‘hats’ of ‘member’ or ‘monadic’ State is not always easy to draw, particularly when government representatives collectively exercise their competence outside the EU institutions.Footnote ¹⁸ But the idea of a functional duplication of representative State organs is not fundamentally problematic. It can be solved by a duplication of legal orders depending on whether States act as monads in the international legal order, or as members in the European legal order.Footnote ¹⁹

The idea of duplicating the citizen component of representation seems at first sight more problematic. It is implicitly conveyed by Article 10(2) TEU when it states that citizens are represented in the European Parliament ‘at Union level’. While it might seem like a pleonasm, since the European Parliament is obviously situated ‘at Union level’, it is meant to emphasize that citizens are also represented at another level, that of their Member State, which forms the other ‘indirect’ component of representation in the Union (see above, Section 13.2.2). This is also the meaning that can be given to Article 9 TEU in fine, reproduced by Article 20 of the Treaty on the Functioning of the European Union (TFEU) and according to which ‘[c]itizenship of the Union shall be additional to and not replace national citizenship’.Footnote ²⁰ Yet, while it may seem conceivable that a complex and abstract entity such as a State be duplicated in some cases, the duplication of the citizen as an individual endowed by definition with a single will faces an obvious objection: how could one and the same citizen be represented at two levels, without risking self-contradiction in the event of a conflict between these levels of representation?

This situation is fairly common. It is the case, for instance, when in the State context, a local representative institution collides with a national representative institution, but the issue is then decided based on a hierarchy of levels of representation articulated around the idea of a single and indivisible people. In the absence of such a single people, multiple citizenship is again problematic, even schizophrenic. An example of such a situation was when British nationals tried to pit their status as European citizens against that of national citizens in order to challenge the Brexit effects and continue to benefit from their rights deriving from their previous European citizenship. Presented with such a claim, the Court of Justice merely established a chronological relationship between the two forms of citizenship, national and European, and held that, after the entry into force of the withdrawal, the applicants could no longer claim their European citizenship since they ‘no longer hold the nationality of a Member State, but that of a third State’.Footnote ²¹ Such a chronological approach is questionable: not only does it obscure the fact that, unlike with third-country nationals, European citizenship had been acquired before being revoked,Footnote ²² but it also introduces a form of hierarchy between citizenships. It is difficult to understand how the exercise of national citizenship can call into question the existence of European citizenship, without assuming a form of superiority of one over the other. In fact, what this case law highlights is the derivative character of EU citizenship in relation to the citizenship of a Member State. This example shows the difficulty of articulating a dual and equal citizenship.

One way to overcome the difficulty of the egalitarian duplication of citizenship is to take into account the political nature of representation. As Jürgen Habermas highlights, the construction of citizens’ political opinions takes place within a public space of deliberation that ‘constitutes the necessary connecting element between the political autonomy of the individual and the formation of the common political will of all citizens of the State’.Footnote ²³ It is therefore necessary to distinguish the ‘raw’ interest of the individual as a member of society (Gesellschaftsbürger) from the final interest of the citizen as a member of the political collectivity (Staatsbürger), which is formed only through a space of collective deliberation. The political preferences expressed by citizens are embedded in communication structures that guide their choices, depending on the preferences of the other participants in the deliberation and the available strategies. However, nothing prevents this public space from being dual (national and European), including with a level of porousness between the two, so that the same citizen will be able to formulate different political demands depending on the space of deliberation in which they participate. Thus, under the condition of a two-level public space sufficiently developed to establish a close connection between the citizens and their representatives, the citizens could very well be, or feel, involved in the definition of the national interest that will be defended by their representatives in the Council, and of the transnational interest defended by the European Parliament. Admittedly, conflicts between the results expressed in these two spaces may arise, but this does not necessarily require reducing the citizen to a single or ultimate deliberative community. On this condition, a dual political citizenship in the same space need not be perceived as completely confrontational.

At this stage, one may formulate a first intermediate conclusion to resolve the aporia of the unity of the represented in the European context. The represented is the citizen, himself conceived as a member of two political communities, which justifies that he is represented at European level both ‘directly’ (via the Parliament) as a member of the European political community, and ‘indirectly’ (via the Council) as a member of the national political community. With such a reading, the democratic failure of one level of representation necessarily impacts on the democratic quality of the other. Far from opposing each other, the two levels of representation are bound to strengthen each other.Footnote ²⁴ The question then arises as to whether the conditions for a duplication of political citizenship in the EU are really met. Which leads to the second difficulty: European political citizenship still seems to be largely correlated to the national public space. In those circumstances, it is difficult to consider its holders to be represented as such.

13.3.1 From Peoples to Citizens

Pursuant to Article 10(2) TEU, citizens are ‘directly’ represented at the Union level by the European Parliament. But what citizenship does the European Parliament represent? A truly autonomous European citizenship, or a derivative, or even ‘indirect’, form of national citizenship? The question is whether, from an institutional and procedural point of view, the parliamentary system of European representation is giving itself the means to achieve the ambition it displays. If an evolution is occurring, it is still in progress.

The establishment of a form of democratic representation of a parliamentary type within the institutional system of the Community was necessary from the outset. It mirrored the creation of international assemblies by numerous treaties after the war (Western European Union, Council of Europe, North Atlantic Treaty Organization (NATO), etc.), but also stemmed from the desire to establish a democratic control over the activities of this independent supranational authority, the High Authority, precursor to the current European Commission. The European Coal and Steel Community (ECSC) already included a Common Assembly. The European Economic Community (EEC) retained the same institutional structure, including an assembly that was intended to reflect the association between the citizens of the Member States and the construction of Europe. This assembly, which took the name ‘European Parliament’ in 1962,Footnote ²⁵ is today the parliamentary institution of the EU.

Article 137 of the Treaty of Rome of 25 March 1957 establishing the EEC provided that the Assembly ‘shall be composed of representatives of the peoples of the States united within the Community’.Footnote ²⁶ This formula set aside the idea of representing a European people in favour of representing national political communities. The demoicratic nature of representation in the European Parliament was further reinforced by the method of appointment initially chosen. This method established an organic link between national and European parliamentarians because Members of the European Parliament (MEPs) were chosen by the national parliaments from among their members in accordance with the procedure laid down by each Member State.Footnote ²⁷ Admittedly, the principle of election by direct universal suffrage had been provided from the outset.Footnote ²⁸ Article 21 of the ECSC Treaty left it to the discretion of each Member State to have members of the national parliament chosen from among its members or to provide for the election of MEPs by direct universal suffrage. Article 138(3) of the EEC Treaty instructed the European Parliament to prepare a draft to that effect, on which the Council had to decide unanimously and which had to be ratified by all the Member States in accordance with their respective constitutional requirements. It was not until 1976 that the principle of the election of MEPs by direct universal suffrage was put into practice.Footnote ²⁹ But direct election did not change the nature or scope of representation. As the French Constitutional Council emphasized, the Act concerning the election of the Assembly of the Communities by direct universal suffrage does not intend to ‘modify the nature of this Assembly, which remains composed of representatives of each of the peoples of these States’ or ‘to create sovereignty’.Footnote ³⁰ The election was organized within the Member States, which were free to determine the holders of the right to vote and stand as a candidate. In principle, they reserved it for their nationals.Footnote ³¹

One might have thought that the introduction by the Maastricht Treaty of EU citizenship (to which is attached the right to vote and stand as a candidate in European elections in the Member State of residence under the same conditions as nationals) was an innovation.Footnote ³² This seemed to be confirmed by the Treaty of Lisbon which finally established a link between European citizenship and representation in the European Parliament. The hitherto immutable formula describing the European Parliament as an assembly consisting of ‘representatives of the peoples of the States brought together in the Community’Footnote ³³ no longer appears. From then on, the Treaty provides that ‘citizens are directly represented, at Union level, in the European Parliament’Footnote ³⁴ and describes MEPs as ‘representatives of the Union’s citizens’.Footnote ³⁵

According to the French constitutional judge, there had been a change. In 2003, the French Constitutional Council found that the creation of eight electoral constituencies instead of a single jurisdiction did not infringe the principles of the indivisibility of the French Republic and of the unity of the French people. According to the French Constitutional Council (Conseil constitutionnel), since the Treaty confers on EU citizens the right to vote and to stand as a candidate in elections to the European Parliament in the State of residence under the same conditions as nationals of that State, ‘Members of the European Parliament elected in France shall be elected as representatives of the citizens of the European Union residing in France’.Footnote ³⁶ This highlights a change, or a potential change, in the nature of the democratic representation assured by the European Parliament. The Parliament would no longer be the representative organ of the national political communities of Member States, but that of a new body politic transcending national political bodies and made up of European citizens. In other words, the European Parliament would have become the organ of expression of a European people which is, if not already there, at least in the making, if we are willing to believe in the performative function of law and legal language and admit that political representation constitutes the represented subject. An additional sign of this change in the nature of representation in the European Parliament would be the severance of the link between the European Parliament and the national parliaments in 2002 by the decision amending the Act concerning the election of the representatives of the European Parliament by direct universal suffrage, which made the national and European parliamentary mandates incompatible (as was already the case in France).Footnote ³⁷ But the text of the treaty is not unequivocal and several elements might question this reading.

13.3.2 A ‘Second-Order’ European Political Citizenship?Footnote ³⁸

First, the national fragmentation of the electorate and the ballot’s territorialization into States persist. The State remains the electoral basis for the European elections, which are organized in each Member State for the representatives elected in that State.Footnote ³⁹

Second, the distribution of seats in the European Parliament is subject to quotas per Member State and does not apply strict demographic proportionality. Indeed, according to Article 14(2) TEU, ‘[r]epresentation of citizens shall be degressively proportional, with a minimum threshold of six members per Member State’, with the specification that ‘[n]o Member State shall be allocated more than ninety-six seats’. This means that ‘the ratio between the population and the number of seats of each Member State … shall vary in relation to their respective populations’, but ‘each Member of the European Parliament from a more populous Member State represents more citizens than each Member from a less populous Member State and, conversely, that the larger the population of a Member State, the greater its entitlement to a large number of seats’.Footnote ⁴⁰ The application of that principle of degressive proportionality results in an imbalance in demographic representation, such that the vote of the citizen of the least populous Member State weighs approximately twelve times more than the vote of the citizen of the most populous State. However, as the German Constitutional Court (Bundesverfassungsgericht) observed,Footnote ⁴¹ if the European Parliament were the representative body of European citizens taken as a political unit, such an imbalance would not be acceptable because it would disregard the democratic principle and its corollary, the equality of citizens in the representation of the vote. It can only be explained and justified because, despite the assertions of the Treaty, the European Parliament remains an institution partly representative of the peoples of the Member States. Therefore, equality between Member States legitimately tempers demographic proportionalityFootnote ⁴² because the allocation of national contingents of MEPs responds not so much, or not only to, the concern to guarantee equal suffrage for citizens, but also to the need to ensure adequate representation of national democracies. This system is capable of reflecting at the level of the European Parliament the balance of political forces in even the smallest Member States.Footnote ⁴³ For a while, the amendment to Article 190(2) Treaty establishing the EEC introduced by the Treaty of Amsterdam emphasized this point: by stating that ‘the number of representatives elected in each Member State must ensure appropriate representation of the peoples of the States brought together in the Community’, it tended, in the minds of its initiators, to ensure that ‘the representation of the smaller Member States should be sufficient to enable the main political currents of a State to have a seat’.Footnote ⁴⁴ In other words, the principle of degressive proportionality which governs the State allocation of MEP mandates is justified by maintaining the division of the European electorate into national electorates whose political pluralism must be represented in the European Parliament.Footnote ⁴⁵

Third, the delimitation of national electorate bodies for the European elections remains within the competence of Member States to determine. EU law does not determine who is entitled to vote and stand as a candidate at elections to the European Parliament. All it obliges Member States to do is enlarging the national electorate to include the European citizens who reside on their territories.Footnote ⁴⁶ As for the conditions for integration, they are freely determined by the State, provided that they are the same for citizens of the Union and for nationals. The Netherlands is thus theoretically entitled to decide that the Overseas Countries and Territories (OCTs) (Netherlands Antilles) are not part of the electoral territory and, therefore, to exclude from the right to vote and to stand as a candidate the European citizens residing in the OCTs. It is only to the extent that Netherlands law recognizes the right to vote and to stand as a candidate to Netherlands nationals residing in a third country that the European principle of equal treatment prohibits them from doing so.Footnote ⁴⁷ Even more so, the United Kingdom has been able to grant the right to vote and to stand as a candidate in European elections to Commonwealth citizens (Gibraltar residents)Footnote ⁴⁸ even though they are not British nationals and, therefore, citizens of the Union, on the ground that, under a national constitutional tradition, they have a ‘close connection’ with the United Kingdom.Footnote ⁴⁹ This does not imply that Commonwealth citizens who are residents of Gibraltar become citizens of the Union and entitled to vote in European elections in another Member State to which they have transferred their residence. The competence of the State to determine the holders of the right to vote and to stand as a candidate in elections to the European Parliament implies an imperfect overlap between the subject which, according to the Treaties, is represented in the European Parliament and the European electorate defined by the States.Footnote ⁵⁰ In principle, the right to vote should be recognized to each of the citizens of the Union.Footnote ⁵¹ This does not mean, however, that this right must be exercised by each of them, since the State is free to set the conditions (age, residence, ineligibilities,Footnote ⁵² incompatibilities, etc.), or even that this right must be reserved exclusively to them. Political unity remains defined nationally. The right of political representation at the Union level ‘is in no way the attribute of belonging to a new political or European community … but rather the mark of integration into a given national community’.Footnote ⁵³ Together with civil, economic, and social rights, the political rights attached to EU citizenship contribute to endowing their holders with the status of ‘quasi-nationals’.Footnote ⁵⁴

If one adds that European elections are still not organized according to a uniform electoral procedure,Footnote ⁵⁵ and, on a political level, the transnational dimension of the public space is not yet sufficiently developed to allow European elections to generate a fully European political agenda,Footnote ⁵⁶ one must admit that, even under the Treaty of Lisbon, representation in the European Parliament continues to be ‘nationally mediated’,Footnote ⁵⁷ and a member of the European Parliament largely remains an emanation of a national community. In sum, the democratic legitimacy of the European Parliament remains more plural than unitary. Symptomatic of this is the fact that the entry into force of the Treaty of Lisbon did not lead the Court of Justice to change its qualification inaugurated in 1980Footnote ⁵⁸ according to which the involvement of the European Parliament in the Union’s decision-making process reflects ‘the fundamental democratic principle that the people should take part in the exercise of power through the intermediary of a representative’.Footnote ⁵⁹

How might it be possible remedy this indexation of European political citizenship on the national public space, this national compartmentalization of the parliamentary representation of European citizens?

13.3.3 Transnationalizing the Public Space and Democratic Representation

Despite the persistent national origin of MEPs, elements of a transnationalization of democratic representation in the European Parliament already exist.

First, the presentation of candidates for MEP under the aegis of European political parties is encouraged. Although the first European political parties were formed in the 1970s with the specific goal of coordinating electoral campaigns in the context of European elections, the Maastricht Treaty enshrined their role in primary law, which is, according to Article 10(4) TEU, to ‘contribute to forming European political awareness and to expressing the will of citizens of the Union’. After the Treaty of Nice included a legal basis to that end,Footnote ⁶⁰ the EU created a European legal status and established a public financing mechanism to promote their emergence and development.Footnote ⁶¹ To benefit from it, the political party must be transnational in scope, that is, represented in at least a quarter of the Member States by MEPs, in national or regional parliaments or in regional assemblies; or have obtained, in at least one quarter of the Member States, at least three per cent of the votes cast in each of those Member States in the last elections to the European Parliament. Moreover, it also must have taken part in the elections to the European Parliament or have publicly expressed its intention to participate in the next elections to the European Parliament. This clearly shows the desire to make European political parties vehicles for the Europeanization of democratic representation. Today, there are nearly fifteen European political parties, formed by grouping together national parties. To give just one example, the European People’s Party, which was created in 1976, today brings together more than eighty political parties, some thirty of which do not come from EU Member States. Yet, the development and constitutional role of these parties remain limited.Footnote ⁶² Deprived of members and partisans (other than through the membership of affiliated national parties) and hampered in the elaboration of global societal projects by the heterogeneity of the conceptions of their member parties and the limitation of the competences of the Union, they hardly are instruments of mobilization and political socialization of European citizens. The electoral manifestos they issue in preparation for the European elections have little echo on the different national stages during the European elections, and MEPs remain preselected by the national parties. At most, they are able to coordinate the positions of the MEPs within the transnational political groups.

As early as 1953, the MEPs decided to sit by political groups and not by national delegations. The transnational dimension of these groups was gradually reinforced by legal texts. In 1997, while the Rules of Procedure of the European Parliament still allowed for the formation of political groups comprising only Members from a single Member State, in an effort to promote the transnational dimension of the political groups, the minimum number of Members required for a group was lowered in proportion to its increase in multinational character.Footnote ⁶³ Today, the Rules of Procedure of the European Parliament of April 2023 only allow the formation of transnational political groups. Article 33 provides that a political group must have at least twenty-three members elected in at least one quarter of the Member States. This structuring according to a transnational political logic is unanimously seen as a sign of ‘fusion’ of European representation, and therefore of the ‘generality’ of the mandate.Footnote ⁶⁴ Far from being the representatives of the national political bodies from which they come, MEPs collectively represent the indivisibility of the European peoples whose collective interest is distinct from the sum of their particular interests. Even the European judge emphasized that the grouping of MEPs by political affinities ‘enables local political particularities to be transcended and promotes the European integration’, and aids in realizing the constitutional objective assigned to European political parties of forming a European consciousness and expressing the political will of European citizens.Footnote ⁶⁵ Representation would thus constitute the represented as an indivisible body politic. Let us beware, however, of giving too much credence to this sign of the mandate’s generality. As evidenced by the fate of the Member of Parliament (MPs) elected in the United Kingdom who, following Brexit, had to leave the European Parliament, the mandate of MEPs cannot be compared to that of a Member of the French National Assembly who, although elected in a constituency, does not represent that constituency, but the nation as a whole.Footnote ⁶⁶ Moreover, the partisan structure of the assembly does not prevent members of the same nationality of a political group from forming national delegations to which, depending on the group, a greater or lesser degree of autonomy is recognized, which allows them, in certain cases, to free themselves from the group voting discipline.Footnote ⁶⁷

The prohibition of binding mandates may also contribute to a transnationalization of representation in the European Parliament despite the national fragmentation of the electorate. Article 4(1) of the Act concerning the election of the representatives of the Assembly by direct universal suffrage provides that ‘[r]epresentatives shall vote on an individual and personal basis. They shall not be bound by any instructions and shall not receive a binding mandate’.Footnote ⁶⁸ The mandate’s representative nature thus frees MEPs from the duty to look after the interests of their national electorate. MEPs are therefore legally in a position to determine the collective interest of European citizens in a complete independence. By way of a parallel, it is worth recalling that the abolition of binding mandates by the deputies of the Third Estate in 1789 in France was intended to allow them to speak on behalf of a collective person rather than merely reflect a will that predated their action.

In the future, the establishment of transnational lists of candidates for the European elections would certainly go a long way towards remedying the national partition of democratic representation in the European Parliament. This proposal is old. Already in 1998, the European Parliament adopted a report which suggested electing 10 per cent of MEPs on transnational lists within the framework of a single electoral constituency.Footnote ⁶⁹ President Macron took up this idea in his speech at the Sorbonne University on 26 September 2017 when he suggested using the quota of departing British MPs for this purpose in view of the 2019 European elections.Footnote ⁷⁰ Since the Treaty of Lisbon, this proposal no longer necessarily requires a revision of the Treaties because Article 14 TEU provides that the European Parliament represents the citizens of the Union. Yet, it nonetheless requires amending the Act concerning the election of the representatives of the European Parliament by direct universal suffrage, which lays down the principle of the distribution of seats by State, and, as such, makes the Member State the electoral territorial basis of European representation. Such an amendment would have to follow a particularly burdensome procedure.Footnote ⁷¹ That is why the proposal did not advance.Footnote ⁷² But it has since been renewed in the European Parliament with the aim of applying it to the 2024 European elections.Footnote ⁷³ It is also one of the forty-nine proposals that emerged from the Conference on the Future of Europe.Footnote ⁷⁴ The system devised by the European Parliament provides for the election of twenty-eight candidates from lists drawn up at the Union level, with voters being called upon to vote twice in the elections: once for a list drawn up within the national framework, and another time for a transnational list. This structure strives to respect the principle (imposed by primary law) of degressive proportionality even for the composition of transnational lists. For this purpose, Member States are classified into three categories, comprising, respectively, the most populous, the medium-sized, and the least populous States. Each list will have to include a succession of groups of three candidates, and each candidate from a group will have to belong to a different category of States. Since the twenty-eight seats are distributed according to the D’Hondt method, each category of States can hope to be represented.

Such a reform would clearly represent a substantial step forward towards the emergence of a European civic community and for the development of a European public sphere. It would strengthen the European dimension of the electoral campaign at a time when it is still too often fragmented by State and focused on national issues. It would also encourage the emergence of European political parties, or European confederations of national political parties. The draft reform further provides that equal treatment and opportunities must be reserved in each Member State for transnational lists as compared to national lists. If, in fulfilment of the European Parliament’s wishes,Footnote ⁷⁵ the establishment of transnational lists were, moreover, coupled with the institutionalization of the currently informal mechanism of the lead candidate (Spitzenkandidat) – which would stand on a transnational list for the votes of all European citizens to defend a political programme – the power of collective self-determination of European citizens and the politicization of the European democratic system, including the democratic dimension of the election of representatives to the European Parliament, would increase significantly.Footnote ⁷⁶

Finally, the development of mechanisms for democratic participation also contributes to the transnationalization of European representation. Activated upstream of decision-making, these participatory mechanisms have the common characteristic of involving individuals in the European decision-making process regardless of nationality. They therefore invite the European representative institutions (the Council and the Parliament) to be more attentive to the expectations of an emerging European society. The expected result is twofold: transnationalizing and democratizing the Union’s representative system.Footnote ⁷⁷ The transnationalization effect of the European legislative process stems in particular from the European citizens’ initiative, which, in order to be admissible, must be signed by at least one million Union citizens from at least a quarter of the Member States.Footnote ⁷⁸ It also derives from the contribution of European social democracy to European representative democracy. Whether European social democracy takes the form of professional representation such as the European Social Committee (whose consultation is binding on the European legislator in the cases provided for by the Treaty), or that of the European social dialogue (which involves the social partners in the development of social standards), it brings not the citizen, but the ‘situated man’ to the political scene, that is to say, the ‘man conditioned by his environment’Footnote ⁷⁹ to give voice to the demands called for by his concrete economic situation.

Promoted by the Treaty of Lisbon,Footnote ⁸⁰ participatory democracy is conceived as a means of correcting ‘the persistent State-national bias of representative democracy in the Union’.Footnote ⁸¹ By striving to bring out and give voice to the European society, it also pushes European democracy to be more representative of that society.

13.4.1 Representation and Representativeness

In recent decades, the demand for democratic representation has become more sensitive to taking into account society’s diversity.Footnote ⁸² It is no longer satisfied with representing the people conceived as a body of citizens equal in rights, but aspires to make representation more faithful to the reality of the body politic, more ‘descriptive’, in particular by improving its inclusive character, or – at least – by remedying its tendency to exclude certain vulnerable groups. The evolution in the vision and perception of society as being composed of multiple identity groups, sometimes antagonistic, often subject to relations of domination and in constant search of emancipation explains this shift to a representation that is more representative. A better representation of these vulnerable groups at the political level would guarantee democratic improvement, by ensuring that political decision-making no longer perpetuates exclusionary phenomena in a systemic way. Yet, by being assigned an objective of inclusion and emancipation of vulnerable groups, representation also changes its nature: it ceases to be seen as mainly procedural or formal and takes on a substantial and more ‘distributive’ dimension of social justice, one of access to positions that determines, by the adoption of laws, the allocation of other valued social goods (education, employment, health, and so on). Such a transformation of the representative ideal raises a challenge that is not unique to European democracy, but also (and primarily) concerns democracy at the national level. Nevertheless, the clash between the change of scale (Europeanization) and the change in nature (substantialization) of democratic representation raises in the EU a particular difficulty which is mainly due to the lack of both institutional and mental structures to signify and politically represent the diversity of a European society under construction. The current modern ideal of representativeness usually develops as a higher and later stage of representation of an already existing national society. In contrast, aiming for a more faithful representation of social diversity on a European scale requires breaking down the elements of a social whole even before it is fully formed and imagines itself as such. As it stands, it is difficult to affirm the existence of a European society sufficiently organized to be re-presented in its diversity. Admittedly, Article 11 TEU includes, among the democratic principles of the EU, the open and transparent dialogue with ‘civil society’.Footnote ⁸³ In practice, however, the dialogue with civil society takes place through the best organized and most powerful actors at the European level, including lobbies, to the point that it is difficult to see it as a true emanation of the European society.Footnote ⁸⁴ This is why, at the European level, representation is torn between two poles: to strive to create a homogeneous whole that does not yet exist, and to reproduce the heterogeneous reality of this inchoate whole.

A move towards a more substantialist approach to European representation should however not be ruled out. Its first translation can be seen in the proposal of Council Regulation to replace the Act concerning the election of the MEPs by direct universal suffrage that was approved by the European Parliament.Footnote ⁸⁵ In addition to allowing for the establishment of transnational lists, the proposal requires respect for gender equality in the determination of candidate lists ‘without infringing the rights of non-binary people’, and provides for the possibility of exemptions from nationally provided thresholds for the allocation of seats for political parties representing recognized national and linguistic minorities.Footnote ⁸⁶ Yet, until the adoption of this reform, the efforts for representativeness continue to come from, and depend on, applicable national electoral laws. Moreover, the issue of gender and minorities obviously does not exhaust the ideal of representativeness of different social groups. This is why the shift towards a greater representativeness of the European society seems to require the introduction of new forms of participation and direct involvement of citizens in European political decision-making.

13.4.2 The Representation and Participation of Society

The desire to represent and involve categorical or sectoral social interests is reflected in the introduction of specific organs and procedures at the Union level, such as the European Social Dialogue or the European citizens’ initiative. Apart from the fact that their transnational dimension contributes to a decompartmentalization of the public space (see above, Section 13.3.3), these participatory mechanisms tend to compensate, to some extent, for one of the weaknesses of the Union’s representative system with regard to at least one of the principles governing ‘representative government’, which is to be accountable during elections and, therefore, responsive to the citizens’ preferences.Footnote ⁸⁷ Despite institutional arrangements and the strengthening of the political link between the Commission and the Parliament, the system is still unable to establish a link between the outcome of European Parliament elections and the choice of those who govern and the adopted policies. It therefore offers voters little opportunity to ‘punish representatives in the polls’.Footnote ⁸⁸ However, these participatory mechanisms giving voice to sectional interests remain confined to a complementary or even secondary role in relation to parliamentary political representation at the European level. The clear mistrust of these participatory mechanisms probably stems from the fact that parliamentary representation at the European level remains recent and fragile, so that the addition of these mechanisms – which potentially challenge majority choice – is perceived as posing a risk of further weakening representation.

As for social partners, negotiation may lead to the adoption of European legislation on social matters. The Single European Act opened the possibility that European social dialogueFootnote ⁸⁹ could lead to the adoption of social rules by means of an agreement,Footnote ⁹⁰ a possibility that was consolidated by subsequent revisions of the founding Treaties and is now regulated by Articles 154 and 155 TFEU. Several directives, relating for instance to working hours or safety at work, have been adopted following trade union negotiations at the European level, giving rise to a form of European social democracy.Footnote ⁹¹ At first, such a procedure was widely promoted, and the jurisprudence even considered it to be an element of ‘the principle of democracy, on which the Union is founded’ because it ensured that, in the absence of participation by the European Parliament in the process of adopting a legislative act, ‘the participation of the people in that process be otherwise assured’.Footnote ⁹² Social democracy was at that time considered an equivalent to parliamentary democracy. Later, however, the social dialogue mechanism not only ran out of steam and struggled to produce new texts, but it was not viewed as triggering any obligation for the Commission to legislatively implement an agreement negotiated by the European social partners.Footnote ⁹³ This institution, whose democratic legitimacy is only indirect, has a right of veto for the sake of protecting its initiative power, which allows it to define the general European interest independently. Admittedly, if the European social agreement had a greater scope, the social partners would have been granted a right to undertake a European legislative initiative, which, for the moment, is not granted to the European Parliament itself.Footnote ⁹⁴ The restriction of the weight of social democracy thus seems justified by the weakness of European parliamentary democracy, which it was supposed to complement, or even compensate for.

The same dilemma applies to the new European citizens’ initiative procedure provided in Article 11(4) TEU. On the one hand, this form of participatory democracy at the European level is supposed to bring the Union’s institutions closer to its citizens, by giving them the opportunity to directly orient European political action. It contributes to the creation of a transnational public space (see above, Section 13.3.3), by stimulating the formation of sectoral demands within European society. This is why the Court of Justice has chosen to develop its potential, by relaxing the material conditions for the admissibility of an initiative,Footnote ⁹⁵ and by obliging the Commission to cooperate more with the initiators.Footnote ⁹⁶ On the other hand, the Court’s case law has upheld the procedure’s main obstacle, which is to leave the Commission free to decide on the follow-up to be given to a citizens’ initiative. The justification for this blocking power lies in the fear of a competition between ‘representative’ democracy and ‘participatory’ democracy,Footnote ⁹⁷ as the latter can only ‘complement’, and not supplant, the former.Footnote ⁹⁸ As the Court of Justice specifies, the interest of the citizens’ initiative procedure remains limited: it is not to impose a political agenda on the representative institutions, but only ‘to initiate debate on policy’.Footnote ⁹⁹ This explains why the European Commission retains its veto. Again, as long as the EU Parliament does not have a right of legislative initiative, there can be no question of recognizing such a right for citizens’ associations.Footnote ¹⁰⁰ This would extend participatory democracy beyond its role of ‘complement[ing]’Footnote ¹⁰¹ the democratic legitimation of European public authority.

It is difficult to assess the sincerity of the argument of a risk of rivalry between the classic form of political representation and the newer form of citizen participation at the European level.Footnote ¹⁰² The real fear may be that of creating a way to challenge the actions of the EU institutions. In practice, many European citizens’ initiatives call into question the political choices of the majority made by the representative institutions.Footnote ¹⁰³ Yet, a living democracy, one that is truly representative of the diverse aspirations of its society, would probably benefit from taking a better account of dissent and giving it a voice so that it can exert a real influence on its representatives, or else it might reinforce the idea that citizen mobilization only bears fruit at the national level.Footnote ¹⁰⁴ To integrate protest, it is necessary to be able to respond to it, by organizing the deliberation of society with itself on a European scale.

13.4.3 Representation and Deliberation in Society

For the first time, a ‘Citizens’ Conference’ was held at the European level in 2021–2022 to outline the prospects for the EU’s institutional and political evolution. The participatory mechanism established by the Conference on the Future of Europe was special in that it emphasized its deliberative dimension. This deliberative turn extends to the European level a paradigm that is spreading within national democracies with the ambition of taking better account of what citizens want, but also, and above all, of focusing on the way in which citizens form their political will, which must be egalitarian and inclusive. The deliberative process matters as much as its outcome, given that deliberative exchanges are supposed to influence the outcomes, by transforming the initial aspirations of the participants who are guided by both individual and collective reason.Footnote ¹⁰⁵ This democratic form, which somehow reconnects with the roots of the democratic model, aims to generate ‘a new type of representatives’.Footnote ¹⁰⁶ It is no longer a question of appointing a governing elite, nor of representing sectoral interests, but of deliberating individuals ‘randomly chosen, for a short period, as ordinary citizens and for specific and limited tasks’.Footnote ¹⁰⁷ The figure of the ‘citizen-representative’ reappears.Footnote ¹⁰⁸ But since these are ‘representatives’ without a mandate or popular responsibility, their democratic legitimacy depends above all on the transparency and quality of the deliberative process in which they participate.Footnote ¹⁰⁹

Yet, this new paradigm generally presupposes the prior existence of a single or ultimate democratic perimeter, within which the deliberative mechanism is instituted.Footnote ¹¹⁰ In addition to the difficulty of organizing deliberations on a ‘large scale’,Footnote ¹¹¹ in the EU there is also the difficulty of organizing them on a ‘double scale’. The initial obstacle of the duality of democratic levels in the EU strikes again. This is why the composition of the Citizens’ Conference on the Future of Europe sought to combine the two dimensions of citizenship, national and European. On the one hand, citizens’ conferences were organized in parallel at both levels: national and European. In the latter case, four panels of 200 citizens from across the Union were formed by taking into account the geographical origin (nationality and urban/rural environment), gender, age, as well as socio-economic background and education level. This guaranteed a certain form of representativeness, especially for young people (one third of participants were young people aged sixteen to twenty-five). On the other hand, the final Citizens’ Conference, which was responsible for summarizing national and European deliberations and adopting the final recommendations, was composed of eighty citizens from the European panels, but also of one citizen from each national panel. That is, 107 citizens initially designated by a random drawing, plus the President of the European Youth Forum. A whole series of other representatives still completed this citizens’ assembly of 108 people, which gave it a plethoric and hybrid composition, half-deliberative and half-representative.Footnote ¹¹² A total of 449 participants adopted forty-nine proposals, which were further divided into more than 300 concrete measures. A digital platform also collected nearly 50,000 contributions to the debate.

It is not easy to measure the real contribution of this type of complex deliberative structure at the European level.Footnote ¹¹³ Although it lasted for more than two years, it remained largely unknown to public opinion and the national media. From this point of view, it does not seem to have contributed much to the transnationalization of public space.Footnote ¹¹⁴ The experience of the Citizens’ Conference could, however, pave the way for other forms of transnational deliberation, such as the convening of thematic Citizens’ Conferences prior to European legislative work. Moreover, unlike the European citizens’ initiative, which was perceived as contesting and competing with parliamentary representation, the experience of the Citizens’ Conference seems to be more congruent. It is interesting to note that the Conference came out in favour of strengthening European parliamentarism, notably by proposing to confer a right of legislative initiative on MEPs or to set up transnational European elections (proposal n°38). Supporters of a strengthening of the ‘direct’ form of democratic representation could thus draw from transnational citizen deliberation a democratic justification for a real overcoming of the national State framework of democracy.Footnote ¹¹⁵ On the other hand, if most of the proposals of the Citizens’ Conference were to go unheeded, which is not to be ruled out in the light of national experiences, a new democratic disappointment will then likely arise.

14.1.1 A Look Back at the Crisis of IO Legitimacy

The recurring theme, in legal doctrine, of the deficit or crisis of legitimacy of IOsFootnote ⁵ calls for certain clarifications and nuances to unveil possible biases.

While normative legitimacy cannot readily be disentangled from the sociological legitimacy of institutions,Footnote ⁶ a first bias – of approximation (or exaggeration) – is detectable in the perception by scholars of a widespread crisis of legitimacy of IOs. When subjected to closer scrutiny, sociological surveys show that the crisis of confidence in or legitimacy of IOs is not as deep-rooted as is suggested;Footnote ⁷ it could have its roots far from the institutional structures of IOs, in the degree of intersubjective trust that prevails in a society,Footnote ⁸ or in IOs’ incapacity to tackle the concrete problems facing populations; and, finally, it does not generate a particularly strong demand for the democratization of IOs,Footnote ⁹ at least in the sense of the institution or ‘re-institution of IOs’Footnote ¹⁰ by the people or peoples they supposedly represent or of making IOs more representative of Member States and infra-State constituencies.

Dramatization can also result from the evaluation of IOs’ legitimacy deficit in terms of normative standards that are not firmly enshrined in positive law, even though they may be there in outline. This is particularly true of the lack of democratic legitimacy attributed to IOs. This grievance is all the more serious given that the real power of IOs, understood here in the Foucauldian sense of their capacity to exert an action on the actions of others, far exceeds their formal power (IOs’ power to prescribe and condition public policies exceeds, for example, the decision-making power attributed to them by their constituent treaty). While a principle subjecting IOs to a requirement of direct or even indirect democratic legitimization would contain an obvious promise of emancipation for peoples and individuals in relation to the inter-statism still prevalent in international society, its existence in positive law remains dubious. Even if the principle of democratic legitimacy could be invoked against their members, which is far from certain, most IOs have been deliberately instituted and have persevered in being on the basis of an exclusively governmental representation of States – and, through them only, of peoples – without regard to the democratic form of State government. This could correspond to a lex specialis.

Focusing on the legitimacy of IOs considered in isolation is another possible bias. Since IOs rarely have the full power to adopt, impose, and implement norms (or policies), the legitimacy of this whole process can only vary with two factors: the balance achieved between international and national bodies and the interplay of the principles of legitimacy operating within the IO and in the Member States. A growing ‘democratic discontinuity’Footnote ¹¹ can undoubtedly be observed between the real seat of decision-making, shifted outside the State, increasingly distant from its original consent and less and less controlled by the people or their elected representatives, and the space of democratic legitimization of power, enclosed within States. This is not to say that the State should be idealized as the ultimate instance of true democratic legitimization. On the one hand, national decision-making procedures, unless specially prescribed (e.g. in the form of an obligation to carry out an impact assessment), are not designed to take into consideration the voices of other States and populations affected by instruments or public policy subject to validation by the people or their representatives. Conversely, the detour via IOs and submission to majority rule in IOs have the advantage of forcing, to a certain extent, national bodies to take into consideration points of view other than strictly national interests or the interests of the most powerful groups at national level.Footnote ¹² On the other hand, while certain questions of general interest can be decided within the framework of the Nation-State, and while certain interdependencies are reversible, others are irreducible, notably in environmental and economic matters (barring a sudden disruption in globalization). The leeway that the people or their representatives give themselves is sometimes illusory, and reactions to unreasonable displays of unilateralism (perhaps legitimized by the command of a majority) are often entirely for form. For these reasons, the relevant scale for constructing and assessing processes of political legitimization is probably neither that of the State alone, nor that of IOs alone, but either that, on a meso-scale, of the whole formed by an (or several) IOs and its (their) members, or that, on a micro-scale, of ‘transverse institutional arrangements’ which associate organs of one or more IOs and organs of Member States in the processes of drawing up and enforcing norms (or policies).Footnote ¹³

Finally, beyond the original and therefore distant legitimization of the IO by the consent of the Member States and debatable legitimization by the output (i.e. legitimization that ignores the political principles of the founding and framing of power), the legitimacy deficit may arise from the discrepancy between the purpose for which the IO is set up and its results, or between its stated goals and the interpretation of its mandate by intergovernmental bodies (set up according to principles that do not, or no longer, provide sufficient legitimacy). Leaving aside the risk of projecting onto IOs values, expectations, or ends that are those of the observer rather than of the members or constituencies underpinning the IO, political reform of IOs for the purpose of re-legitimizing these bodies should indeed envisage the revision of their goals, functions, competences, and powers, the determination of the normative constraints weighing on them in view of their nature (notably in terms of the protection of human rights), as well as the redefinition of the conditions under which public affairs (or questions of international public interest) are brought up for discussion, the related instruments deliberated upon, adopted, and monitored, and their authors held to account. Disregarding a critical discussion of IOs’ future purposes and obligations is nothing more than a convenient fiction in this academic exercise.

Under the benefit of this delineation, it appears that legal doctrine, whether or not it tends towards a radical critique of the current IO model, relates their legitimacy deficit to several of their institutional features:

1. (1) unequal representation of Member States: the strictly egalitarian representation of States does not, in itself, ensure that less powerful States make their voices effectively heard in IOs; the differentiation of representation rights in IOs (in the form of privileges such as a permanent seat with a veto, or a weighting of votes, or eligibility criteria in restricted bodies) feeds the suspicion that IOs are designed to prolong relations of domination: ultimately, the distortion between, on the one hand, the relatively fixed distribution of seats and votes and, on the other, the current redistribution of power or competences in the international system reproduces and amplifies inequalities rooted in the experience of colonization;

2. (2) the ‘disordered plurality’Footnote ¹⁴ that has governed participation in standard-setting processes in IOs since the 1990s: the attempt to make IOs more ‘democratic’ or ‘inclusive’ promotes a degree of blurring between the formal prerogatives of IO bodies (intergovernmental or otherwise) and the power conquered by non-State actorsFootnote ¹⁵ (NSAs) over certain decision-making processes; conversely, although NSAs are most often confined to the margins of bodies and selected on the basis of ill-defined or dysfunctional procedures or principles, their presence, according to certain institutional and doctrinal discourses, (re-)legitimizes IOs’ action;

3. (3) the monopolization of power within the institution by government representatives or by experts lacking representative status;

4. (4) the preference for a logic of depoliticization of the issues brought before the IO, to the detriment of a logic of confrontation pitting States and, possibly, social groups against each other.Footnote ¹⁶

Ultimately, the cumulative effect of these discrepancies would be to widen the gap between the degree of power that certain IOs can exercise over their members and over the living conditions of populations or private individuals, and the legitimacy of that power.

14.1.2 A Look Back at Some Responses to the Legitimacy Deficit Through the Democratization of IOs

In response, a number of internationalists have come up with ambitious ideas on the updating of constitutionalist approaches to IOs,Footnote ¹⁷ the advent of a ‘cosmopolitical transition law’Footnote ¹⁸ or the transformation of IOs into democratic institutions.Footnote ¹⁹ Whether idealistic or explanatory or ‘non-ideal’ theories, these proposals are centred on the principle of equality among States, peoples, or individuals (depending on the perspective) and on the right to equal participation and consideration of all those who are or could be affected by the exercise of power by these institutions. In a ‘non-ideal’ theory of ‘international democratic representation’, Samantha Besson argues that IOs must be ‘re-instituted’ as representatives of peoples, without, however, ousting either States or their representatives. Without waiting for this constitutive moment, the potential of a ‘multiple representation system’ dominated by the principle of democratic legitimacy should be explored in and through IOs.Footnote ²⁰

While lacking the space here to enter into the detailed discussion that these proposals deserve, it is worth pausing to consider the idea that IOs are subject, or subjectable, to the principle of democratic legitimacy. However alive democracy may be as a practice or an experience,Footnote ²¹ however flexible the concept, in the current state of international law, this principle can only be found, unadulterated, in IOs that follow the model of the club of States or pursue an integration project, provided that their Member States all share the same conception of democracy and make it a condition of (initial or ongoing) participation in the IO.Footnote ²² It remains alien to the institutions with a universal vocation (hereafter: universal IOs) on which our attention will focus.Footnote ²³

Three considerations drawn from the current state of law and institutions underpin this objection. First, the principle of democratic legitimacy is, at best, emerging in international law at the universal level, and suffers from profound divergences as to what constitutes democracy.Footnote ²⁴ Second, its transposition into the sphere of universal IOs can accommodate neither the omnipresence of the State, interposed before peoples, communities, and individuals, nor the presence, in numbers, of non-democratic States.Footnote ²⁵ Third, the regime of universal IOs is the product of a ‘mix of egalitarian and hierarchical arrangements’:Footnote ²⁶ equality among States is a founding fiction that makes it possible to institutionalize inequalities among members (and therefore among peoples or individuals); thus, this regime is a hybrid of unequal representation by States and participation by NSAs, that may be viewed as ‘drowned’ in a multitude of processes and statuses. Unequal representation or participation is thus embedded, to varying degrees, in the very structure of IOs.Footnote ²⁷ But after all, it may be said, no political form is set in stone, and it is conceivable that IOs could mutate: from interstate bodies instituted with a view to fulfilling a mandate conferred by States, they could become democratic or cosmopolitical bodies forging decisions in the general interest at the universal level; in a less radical variant of the mutation, they could serve as vectors for a more democratic representation of peoples and individuals.Footnote ²⁸

But a ‘non-ideal’ theory of international democratic representation would then come up against another set of objections, this time drawn from the contemporary political context. Unless we project ourselves towards the utopia of a new kind of IO constituted without or against States, the balance of power and the political mindsets of States are decisive, since institutional reforms and, a fortiori, a redesign or refoundation of IOs presuppose, in one way or another, an agreement among States. However, the current political climate does not seem conducive to the deployment of democratization of IOs or democratic representation in and by IOs.Footnote ²⁹ On the one hand, the beneficiaries of the current institutional arrangements, including some of the most democratic StatesFootnote ³⁰ (sometimes exposed to a phenomenon of democratic backsliding), are a priori more interested in fine-tuning than in the complete dismantling of these arrangements.Footnote ³¹ On the other hand, the challenge to the established institutional order is not led only by States in a position of institutional inferiority or of dependence on IOs, all supposedly eager to continue ‘opening up’ the IOs until they are fully democratized. It is also vigorously opposed by openly ‘illiberal’ States (to use an all-encompassing category), whose project is to confine IOs to the representation and protection of States alone, as sovereigns over their territory. The lack of consensus and the paradoxical distribution of votes on a series of United Nations General Assembly (UNGA) resolutions on a Democratic and Equitable International Order (DEIO) reveal the depth of divisions in international society over the very notion of democracy and the desirability of transposing it into the international institutional system.Footnote ³² While the DEIO is pledged to ‘[t]he right to equitable participation of all, without any discrimination, in domestic and global decision-making’,Footnote ³³ some fifty Western and allied States are opposed to it, while the sponsors of the latest resolution (2022) are drawn from among the most illiberal and authoritarian States, including Russia and China, which are engaged in a coordinated effort to subvert concepts borrowed from Western liberal-inspired political rhetoric and redefine the modalities of multilateralism, in a more oligarchic than democratic sense.Footnote ³⁴ To attempt to rush this momentum by formulating reform proposals based entirely on democracy is to run the risk of devitalizing or subverting the concept in the international political arena, or even of creating a formidable ‘paradox of consequences’ (M. Weber).Footnote ³⁵ Indeed, IOs have once again become an arena for ideological confrontations over the foundations and limits of power: States taking political control of them, either as a result of a shift in power, or as a result of a redistribution of seats and votes dictated by the imperative of true equality between States, peoples, or individuals, will be able to impose their conception of ‘international democracy’ on fragile or hesitant States, and protect themselves against the assertion of a universal and robust principle of democratic legitimacy.

For all these reasons, ‘democratic representation’ seems to have no explanatory power for the current structure and operation of IOs (which bring together States and NSAs) and a weak justificatory value for upcoming political reforms of universal IOs. Nothing in this observation precludes reflection on the hypothetical conditions for potential or ideal political models that would transcend national frameworks to guarantee public participation whenever issues of international or transnational interest are at stake. There may even be an urgent need to do so,Footnote ³⁶ given the widening gap between existing political models – including representative democracy – and the challenges we face, particularly environmental ones. But if, in the current context, reflection tends towards the prescription of remedies to the legitimacy deficit of existing universal IOs,Footnote ³⁷ it cannot follow the path of ‘democratization’ of IOs that remain interstate, largely open to non-democratic States, and permeable to manipulations of the idea of democracy.

14.1.3 Searching for an Alternative Route

The alternative approach proposed here takes as its starting point today’s universal IOs and the under-exploited potential of positive international law, while remaining guided by the necessity to preserve the basic conditions and virtues of democracy in the State.Footnote ³⁸ This requires some degree of methodological and political cautiousness. As a preliminary, no-harm rules can be formulated. The first principle is that IO reform should be designed in such a way that IOs do not become, on the pretext of equal representation (of States, peoples, individuals), the crucible of a political project to annihilate the democratic legitimization of power or of organized resistance either to the crystallization of a general principle of democratic legitimacy in a national or international framework or to a more inclusive deliberation within IOs. The second principle, which is also negative, is that the political reform of IOs should not prejudice the regime of political entities that have a democratic regime or are converting to one, since the democratic dynamic can only be expressed, for the time being, within a national or integrative framework (with the limits pointed out in this chapter). In addition, since we must start from what already exists, one fact stands out: universal IOs do not make their Member States’ participation conditional on the existence of an unequivocally defined democratic regime. This absence of democratic conditionality is obviously the sine qua non for maintaining a framework conducive to the indispensable cooperation among, potentially, all States, from the least to the most democratic, that we hold to be indispensable.

Building thereon, we propose to define, positively, principles of political reform of IOs that do not relate their legitimacy either solely to the consent initially given by the State, or to a hypothetical democratic representation in and by IOs. This reflection is rooted in the matrix of IOs: multilateralism. Whatever the degree of concordance between States’ political values, multilateralism on a universal scale is defined and justified as a principle articulated to values, a method of cooperation, and a space for apprehending objects from a multitude of sides. In this way, these objects become truly common and can, if need be, give rise to common questions and answers.Footnote ³⁹

However, and this is what this contribution will focus on, as States and their representatives are not in a position on their own to identify objects that transcend their interests and go beyond their borders, the space of international deliberation needs to be restructured to accommodate, in bodies that have yet to be created for the most part, voices other than those of States, in a way that makes it possible to clearly identify the share of each voice in an enlarged, pluralist multilateral deliberation. In concrete terms, the challenge is to identify why, and under what conditions, new subsidiary bodies should and could contribute to:

1. (1) improving the quality of deliberation within IOs, by ensuring that voices other than those of States or a few ‘stakeholders’ are heard;

2. (2) creating fora for coordinated discussion of projects (instruments and actions) or of the track record of IOs with related, complementary or overlapping skills (better deliberation among IOs);

3. (3) or promoting accountability (better deliberation about IOs).

‘Reinvigorating multilateralism’, ‘constructing international public space’ within the IOs themselves,Footnote ⁴⁰ this is the contribution that can be expected, on a theoretical level, from a ‘reconceptualization’ of the IO’s functions (Section 14.2) and, on a more practical level, from the creation of new subsidiary bodies with consultative functions alongside intergovernmental bodies, welcoming participants from civil society and others (referred to here as NSAs) (Section 14.3).

14.2.1 Rethinking the Functions of IOs and Their Reform

Much has been written in international law on the limits of a functionalist approach to IOs, weakened by ‘blind spots’ (the rights of third parties and human rights) and tainted by biases inherent in the principle-agent theory, the fictitious nature of their political neutrality, or the fragmentation of competences.Footnote ⁴¹ Our intent here is not to fuel a plethoric doctrinal debate that has helped to renew our understanding of IOs, and to reevaluate ‘functional necessity’. However, the relevance of these criticisms does not render the notion of function useless in understanding IOs, since they remain institutions established for certain purposes or with a view to the fulfilment of certain functions which the Member States agree to exercise in common or through them, or to consider as falling solely within the remit of an IO.Footnote ⁴² Goals, functions, competences, and attributed powers can be said to form together the mandate of the IO.Footnote ⁴³ If we are willing to accept this constitutive feature of existing IOs, legitimization through functions remains inherent in them. This is not to say, of course, that it is or should be exclusive of any other principle of legitimacy, nor that it is never misused. On the contrary, it can only be maintained under certain conditions.

However diverse universal IOs may be, unless they have strictly operational or jurisdictional functions, they are invested with ‘metafunctions’ that override their respective mandates. The historical metafunction of this type of IO is undoubtedly to provide a framework for the coexistence of States that are always likely to compete for power, influence, and control over space and resources, a framework that relatively pacifies their relations, at the cost of inequalities in member representation (notably to keep the most powerful States in the institutional system). But there are other metafunctions which today constitute the ultimate raison d’être of universal IOs:

1. (1) opening up spaces for confrontation and deliberation;Footnote ⁴⁴

2. (2) defending universalizable values translated into universal legal principles and common interests;

3. (3) understanding complex interdependencies that transcend national borders;

4. (4) contemplating the long-term, since they are disconnected from national political fluctuations which are not themselves synchronous;

5. (5) and, perhaps, taking into account planetary boundaries.Footnote ⁴⁵

The ‘forum’ function is the ‘only utility that is not contested (at least in the plenary bodies)’.Footnote ⁴⁶

To put it more succinctly, the ultimate function of universal IOs is to constitute the perennial framework for inter- and transnational deliberation on common objects defined by a space, a temporality, and constraints that do not coincide with the coordinates of deliberation in any other public institution at State or sub-State level.

14.2.2 Rethinking the Functional Legitimacy of IOs Based on Their ‘Deliberation’ Metafunction

‘Deliberation’ might seem at odds with the mundane understanding of States’ interplay in/through IOs, supposedly based on bargaining (under threat sometimes) and negotiating. Negotiation is defined, negatively, as ‘neither the result of an aggregate of individual choices (like voting), nor that of a single decision taken by a hierarchical authority’Footnote ⁴⁷ and, positively, as a mode of ‘joint decision-making’ among parties who, having divergent interests, seek a mutually acceptable solution.Footnote ⁴⁸ By contrast, deliberation is ‘[e]xamining, weighing up all the elements of a question with other people, or possibly oneself, before making a decision, in order to reach a conclusion’.Footnote ⁴⁹ Deliberation, then, may or may not lead to the unilateral adoption of an act, which may or may not be legally binding, depending on the features of the law-making, evaluation, or control procedure considered. On closer scrutiny, IOs are indeed the seat of deliberation and voting, both of them permeated by transactions.

It remains to be explained how the concept of deliberation can be transposed to IO theory without contradicting the assertion that democracy is not a relevant concept for universal IOs. It is true that, in political theory, deliberation has often been closely associated with democracy, right up to the construction of a model of ‘deliberative democracy’ according to which ‘the norm is legitimate only if it is based on public reasons resulting from an inclusive and equitable deliberative process in which all citizens can participate and in which they are led to cooperate freely’.Footnote ⁵⁰ Yet, deliberation is not a concept captive to that of democracy, and the conditions for good or better deliberation do not necessarily exclude all forms of transaction – not even in a deliberative theory of democracy. Deliberation can also be articulated with ‘non-deliberative procedures’, such as voting. A more comprehensive approach envisions that participants might have both the common good and their ‘self-interests’ in sight, that they might resort to ‘non-coercive forms of negotiation’ and, ultimately, that lingering divergences might usher in the ‘clarification’ and ‘structuring’ of conflict instead of consensus.Footnote ⁵¹ Conceived of as an alternative to ideal ‘deliberative democracy’, this expanded conception of democratic deliberation can in fact be transposed to non-democratic universal IOs. After all, general principles binding members of IOs are not so remote from the prerequisites for ‘mutual justification’ as to be the ‘central task’ of deliberation between citizens: ‘mutual respect, equality, reciprocity, mutual justification, the search for fairness, and the absence of coercive power’.Footnote ⁵² What is more, IO members have no less interest in arguing than in bargaining.Footnote ⁵³

Finally, decisions by IOs (even if not unanimous) can be considered legitimate through the combination of the legitimacy derived from the State’s initial consent, from the IO’s metafunction (deliberation) and specific functions, from the respect for its particular mandate (subject to evolving interpretation), and ultimately from the fairness of the deliberative process.

14.2.3 Rethinking Pluralism in IOs Beyond the Requirement for Democratic Representation

When one focuses on democratic representation in IOs, a series of a priori requirements, in particular equality of participation in decision-making bodies and procedures, or representativeness, are to be imposed on the constitution of IO bodies. But when the purpose is to define the conditions for good or better deliberation in IOs, the yardstick to determine which voices should have which share – given that different perspectives should be confronted but that each is imbued with a potential bias towards unilateralism – is to be found in their metafunctions and specific functions. These conditions prefigure and configure the exercise of each IO’s own mandate.

Several avenues can be explored, simultaneously or not, to (re)create or maintain a space for interstate and transnational deliberation that is as open as possible, and therefore pluralist, enabling the settlement of public affairs and the assumption of functions of international public interest (within the mandate of each IO).

Hence, the rebalancing of State representation within IOs. While this is impossible to justify in terms of democratic requirements,Footnote ⁵⁴ it can be done, classically, by updating the representation of the collectivity of Member States, per se (egalitarian or specular representation) or with regard to the functions of the IO (a potentially differentiated representation) or, in the perspective defended here, with regard to the need for pluralization of the voices heard in IOs. Institutional arrangements are available, such as the enlargement of restricted bodies to currently under-represented States, the rotation of seats on restricted bodies, or the maintenance of the ‘one State one vote’ rule in combination with complex decision-making rules that protect all members against automatic majorities or institutionalized unilateralism, and force Member States, to either compromise or display outright opposition. The IO’s capacity to take decisions may suffer as a result, but this is the price to pay if the greatest number of States are to retain a genuine interest in international cooperation. In order to re-establish a certain continuity in the legitimization of the power shared between States and IOs, the participation of people’s representatives, designated within a national framework, in the definition of positions taken by the Member States in the IO and in the adoption of IO instruments may involve the systematic inclusion of parliamentarians in national delegations, or the creation of (inter)parliamentary bodies.Footnote ⁵⁵ This will not result in democratizing decision-making in IOs since not all representatives of the people will be freely appointed. Nonetheless, deliberation in IOs could thus be more pluralistic.

The expansion of multilateral deliberation should not, however, benefit only Member States and their parliaments. Indeed, universal IOs are already marked by the strong presence of NSAs. It was enough for Member States to support or tolerate the opening up of universal IOs to entities other than themselves for multiparty ‘representation’ or participation to take root. It is not ‘democratic’ (despite the institutional rhetoric) but could gain in legitimacy and legitimizing force if justified and adjusted to the requirements of better deliberation. This presence should be amplified, but rationalized, for reasons of principle: the object of deliberation in IOs cannot be apprehended solely through the confrontation or a sum of statist points of view because States are themselves subject to limitations arising from their difficulty in projecting themselves into the long-term,Footnote ⁵⁶ in apprehending the negative cross-border externalities of public policies,Footnote ⁵⁷ and, even in the most democratic States, in hearing the plurality of all relevant voices on a subject.

Deliberation institutionalized in the IO can and must therefore in abstracto accommodate persons with a representative mandate associated with IO membership, but also experts (subject to robust and controlled rules of independence) without a representative mandate,Footnote ⁵⁸ as well as persons giving voice either to natural entities (with or without legal personality) or making their point of view heard or to groups that find no satisfactory political expression in any State (e.g. refugees or migrants), to peoples without a (universally recognized) State, to marginalized groups, and so on.Footnote ⁵⁹ Deliberation can also be enriched by the ad hoc direct participation of particularly interested categories of people, in the form of petitions or ‘pure’ or hybridFootnote ⁶⁰ transnational citizens’ conventions. In fact, deliberation can be adapted, according to the functional specificities of each IO in terms of designating the categories of subjects, interests, or perspectives that are granted an institutional role and/or status, and defining the procedural rights of each. The proposal to transform universal IOs from fora in which a ‘polylateral diplomacy’ with multiple channels or voices already unfoldsFootnote ⁶¹ into restructured and structuring spaces for inter- and transnational deliberation is clearly liberal in inspiration: a quest for greater inclusivity and greater rationality in the product of deliberation is at stake.Footnote ⁶²

Focusing on IOs, a good or better deliberation will be one that:

1. (1) may lead to varying degrees of change in positions initially defined from a necessarily one-sided perspective or on the basis of partial information;

2. (2) will lead, at best, to a reasoned outcome, taking the form of a majority decision or no decision at all, or at least to a ‘reasonable dissensus’;Footnote ⁶³

3. (3) affords comparison between the final outcome and the particular positions of each party at each stage, on the one hand, and between the arguments exchanged, on the other.

The concrete determination of the parameters of this space for deliberation takes place by descending from the level of the metafunction to the functions and powers of the IO and the bodies designated to carry them out.

14.3.1 Reshaping the Institutional Architecture

Institutional architecture follows basic patterns that spread by isomorphism. The dominant, strictly intergovernmental models are still so powerful that the International Renewable Energy Agency (IRENA) was set up in 2009 without a body designed to accommodate NSAs, despite the sensitivity of its field of competence to actors engaged in the ecological transition.Footnote ⁷⁰ However, the functioning of IOs already depends on the possibility granted to actors other than members to participate in the deliberation process (with the rights that are those of an observer, for example) and on the creation of additional, so-called subsidiary bodies. These bodies may themselves be intergovernmental, but they can also accommodate people with no representative mandate, as is the case with expert bodies. In this way, they offer a particularly interesting institutional format for remedying, in part, the growing dissociation between the purpose of IOs, which is no longer strictly interstate, and their structure, dominated by intergovernmental bodies with a contested balance.

Designed for one, two, or more of the IO’s intergovernmental bodies, and open to actors other than States or IOs (grouped here under the generic name of NSAs) to enhance their deliberation, consultative bodies would be distinct from multi-stakeholder fora, networks, or platforms set up by an IO with the explicit aim of helping it fulfil one of its prime objectives, under the control of the integrated administrative body.Footnote ⁷¹ Functionally, they would be closer to certain recurring multi-stakeholder fora which, like the Youth Forum of the Economic and Social Council (ECOSOC), can lead to NSAs adopting their own positions.Footnote ⁷² Their institutional status, however, would be that of IO bodies with attributions formally articulated to those of the main bodies.Footnote ⁷³ This is an essential point. The UN Permanent Forum on Indigenous Issues created by the ECOSOC in 2000 as a subsidiary organ with representatives of indigenous peoples sitting in their personal capacity as ‘independent experts’ could inspire further practice.Footnote ⁷⁴

14.3.2 Participation of Non-State Actors in Consultative Subsidiary Bodies

Defining rules for participation is key for adequately enlarging and improving deliberation within IOs.

15.5.1 Challenges of Democratic Representation in and by IOs

Contributors are raising a number of questions in relation to democratic representation in and by IOs. Some of them are quite specific, others have a broader scope and mean to go to the core of the argument. Without the ambition of being exhaustive, I am mentioning four of these challenges below.

15.5.2 Alternative Sources of Democratic Legitimacy of IOs?

Several contributors have put forward the idea that representation may not be the sole source of legitimacy in IOs. ‘Representation’ itself, as we have seen above, may be interpreted in a looser sense of ‘representativity’, with the concepts of ‘descriptive’, ‘mimetic’, ‘statistic’, or even ‘altruistic’ representation. Besson and Martí observe in this regard that these forms of representation ‘purport to represent pre-identified traits or interests to be promoted (functional output) or, at least, knowledge or expertise in how to do so (functional input)’.Footnote ⁶² But in any case, they cannot be an equivalent or constitute ‘democratic representation’ in the strict sense, that implies ‘ultimate and effective popular control’. Still, in the practice of IOs, this loose sense of representation is often considered as a superior value when it comes to participation of private actors than ‘democratic representation’.

In his chapter, Pierre Rosanvallon made eloquently the point that the most important social movements today are not representative movements, in the strict sense.Footnote ⁶³ Marie-Clotilde Runavot on her part showed that what was valued in the participation of domestic parliamentarians in IOs (with the Inter-Parliamentary Union) was not so much their representative status, but rather their substantive contributions.Footnote ⁶⁴ And Évelyne Lagrange made the point that linking democracy to representation (strictly conceived) does not give a clear basis for the participation of actors that are not ‘representative’ but are nonetheless legitimate participants. And she argues that ‘decisions by IOs (even if not unanimous) can be considered legitimate through the combination of the legitimacy derived from the State’s initial consent, from the IO’s metafunction (deliberation) and specific functions, from the respect for its particular mandate (subject to evolving interpretation), and ultimately from the fairness of the deliberative process’.Footnote ⁶⁵

Indeed, there are other forms of legitimacies of power than ‘democratic legitimacy’ in the strict sense of a representative democracy:

Legitimacy based on knowledge, that is epistemic legitimacy or what Évelyne Lagrange called ‘epistocratic’ legitimacy,Footnote ⁶⁶ is taking a growing part in the work of IOs, with experts, including scientific experts, or even specialized NGOs, that bring to an IO knowledge and information that it cannot acquire by itself.

Legitimacy based on influence and/or financial impact: one just has to think of the role played by the Bill & Melinda Gates foundation in the World Health Organization (WHO) and more generally in global health issues.Footnote ⁶⁷

Watchdog legitimacy, that is the legitimacy of those who are holding powers to account, like investigative journalists, human rights defenders, and whistle-blowers … that is very often persons who do not represent anybody but themselves but whose legitimacy is based on the fact that they ‘speak truth to power’.

One may wonder however whether these types of legitimacies have something to do with ‘democratic legitimacies’. Here Melissa J. Durkee makes the point that the question of participation of private actors in IOs reveals a tension between what she calls ‘a logic of representation’ and a ‘logic of expedience’. Legitimacy in the first logic is based on the representativity of the private actors, whereas in the second logic, the additional value these actors can bring to realize the goal of the IOs is a primary consideration. She sees the first logic as flowing from a ‘deontological perspective rather than a consequentialist or teleological one. That is, democratic self-determination and interest representation are inherent goods. Decisions made, rules developed, and actions taken through a legitimate process are right and correct because they are products of that process’.Footnote ⁶⁸ Conversely, ‘[t]he logic of expedience is defined by its embrace of consequentialist and teleological reasoning. An outcome is right and correct when it advances the mission of an organization. This position values efficiency, problem solving, and progress on accomplishing public goals, whether they be public health, climate stabilization, sustainable development, etc.’Footnote ⁶⁹

The point here is that it seems that in some instances those logics seem irreconcilable: IOs would have to choose between the two logics. Either having democratic representatives, or choosing the participants for other sorts of legitimacies such as ‘epistocratic’, ‘watchdog’ or influence.

Lagrange tends to relativize this opposition in arguing that democracy is, after all, legitimizing the exercise of power and that, to this end, representation may not be the exclusive or even the best source of legitimacy. Representation, she submits:

In other words, she argues that democratic representation should not be seen as the sole basis for democratic legitimation, quite the contrary: ‘epistocratic’ legitimacy should be considered as one of the main bases of legitimization, from the moment that it is placed in a pluralist environment, and that it is linked and combined with other principles of legitimacy, including democratic representation.Footnote ⁷¹ I would add to this that the justification for participation is very often a question of point of view: IOs may see participation of actors through a logic of expedience, whereas these actors may understand their role as participating to a logic of representation and, beyond, of democratization of IOs.

15.5.3 Current Processes and Models of Global Governance

Besson and Martí are proposing a MIRS in a ‘systematic and ordered approach’, rather than a ‘radically pluralistic’ one.Footnote ⁷² Even though the authors are certainly right in pointing out that current systems of participation are not in line with a strict conception of democratic representation, several contributions suggest that they are not so disorganized. In fact, most of participatory or consultative arrangements in IOs, for actors other than States, are following similar principles. These maybe more akin to a logic of expedience than a logic of participation, but at the same time we have seen above that the two are not necessarily incompatible and that, rather, IOs are combining the two to give way to a combination of legitimacies, which, in the end, may give an appropriate foundation to the exercise of power.

Yves Sintomer has suggested during the conference that the concept of ‘governance’ may be looked at when trying to understand and describe those types of processes. For sure, ‘governance’ is a word saturated with controversies. Although Jacques Chevallier,Footnote ⁷³ but also André-Jean ArnaudFootnote ⁷⁴ have tried to overcome this confusion and get the most of a concept which, after all, may well be extremely useful to advance knowledge about how power is exercised not only in domestic societies but also at the international or global level.

Chevallier points out that the State in the post-modern age is faced with problems of ‘governability’, insofar as it has to deal with ‘demanding citizens who are no longer satisfied with the principle of delegation inherent in the representative system but want to have a say in collective choices’ – which leads to a search for ‘new methods, different from traditional government techniques’.Footnote ⁷⁵ These ‘new methods’ are generally subsumed under the term ‘governance’.

Chevallier continues:

Even though not all contributors have used the word ‘governance’, much of what they describe fits with Chevallier’s account of the concept.

Global governance means that authority is dispersed between different players, at different levels and in multiple ‘sites’, which tend to form a ‘deliberative system’ through interaction between these ‘interested parties’. This is not about a right to command, in the classical sense of sovereignty of the State: ‘governing the globe’Footnote ⁷⁷ is not of the essence of imperium but more of auctoritas. Niko Krisch uses the idea of ‘liquid authority’ to explain that type of power that circulates among actors, and is neither founded on force or persuasion, but rather on its capacity to create deference on the side of other actors.Footnote ⁷⁸ In that, governance is different from government. More often, it is not even much about making legal norms, but rather public policies that are action-oriented. Such governance processes can be observed when IOs are entrusted or assume the responsibility of defining global public policies on global issues (such as sustainable development) involving multiple partners beyond the States at multiple levels – international but also national, regional, or local.

Participation in governance in this regard means participation in the exercise of public authority; it is a way in which ‘legitimate’ actors participate in the exercise of ‘fluid’ public authority jointly with the State and IOs. Increasingly used in IOs and governance processes, the notion of ‘partnership’ reflects this idea of shared power.Footnote ⁷⁹ And this sharing takes place not because of a decision taken by an original exclusive holder of power, but because the actors involved – the stakeholders – are themselves originally custodians of a significant portion of a circulating power – a ‘liquid’ power that cannot be completely ‘concentrated’, ‘solidified’ to be concentrated in a single subject.

In this context, following Rosanvallon, ‘representativeness can even come to be perceived as a burden, an insoluble constraint’.Footnote ⁸⁰ This is very well illustrated in Marieke Louis’ chapter, about ‘classical’ representative organizations like IOE being much less at ease than the International Chamber of Commerce in the context of post-2008 financial crisis economic negotiations.Footnote ⁸¹

Another key concept to understand governance is the concept of ‘orchestration’, as demonstrated in Terry Macdonald’s chapter on the Global Compact on Refugees. What Macdonald shows is that ‘orchestration’ may have the effect of putting some groups in contact and inciting them to coordinate their action and/or to act through a loose network organization based on the sense of community or shared experience. As Macdonald points out, ‘[t]hrough so doing, IO orchestration activities can potentially support not only the functional efficacy of global governance practices, but also their democratic legitimacy’.Footnote ⁸² By ‘orchestrating’, that is convening, endorsing, but also assisting, IOs may encourage ‘mimetic’ or ‘statistic’ representatives to constitute themselves in groups that may pretend to represent their constituencies. Aspirational democratic representatives would in general take the shape of coalitions of associations, rather than a unique and unified organization. And that is because most of the constituencies – unless maybe very limited ones or localized ones – would in general produce a number of would-be representative organizations, corresponding to various groups within them, as fragmented as they are along political, social, cultural, ideological, or other lines. This pluralism within constituencies is in fact itself consubstantial with democracy. Far from being undemocratic, this scattered and multiple representation by various non-representative (in the strict sense) actors may in fact enhance democracy in IOs.