The European Union as a demoicracy: Really a third way?

Introduction

Should the EU be a full-blown federal union or merely an intergovernmental forum of fully sovereign states? This is a central question in EU scholarship across several disciplines. Similarly central is the question around the EU’s alleged ‘democratic deficit’, namely its imperfect democratic accountability to those to whom its rules and regulations apply. The two debates are largely intertwined, with federalists and intergovernmentalists often taking opposite views on what, if anything, is democratically wanting about the EU.

Recently, however, a still fairly self-contained (yet rapidly expanding) literature has been arguing that the choice between a Europe of states and a European ‘Superstate’ (Morgan, 2007) is based on a false dichotomy, and that this is due to a misunderstanding about the kind of democratic ideal that the EU should embody. ¹ According to such a body of literature, there exists a viable alternative to both institutional models. What is more, this alternative not only constitutes a desirable ideal in and of itself but is also a more faithful reconstruction of the original motivation behind the project of European integration as well as of its ongoing aspiration. The idea, in a nutshell, is the following: the EU is – and its performance should be normatively assessed as that of – a demoicracy, namely a ‘Union of peoples [not a people] who govern together, but not as one’ (Nicolaïdis, 2013: 351). The demoi of Europe recognise that they unavoidably affect one another’s democratic health in problematic ways, and that this generates reciprocal obligations. They therefore establish a union to guarantee their joint and reciprocal freedom qua demoi – which most demoicrats cash out, in a republican spirit, as non-domination. This is much more than intergovernmental multilateralism, because the demoi genuinely govern together on the matters that concern their reciprocal freedom. However, the institutional model that stems from this must be faithful to what the demoicratic union is a union of. If the union aims at protecting the freedom of the different (and inherently diverse) European demoi, then it cannot permissibly do so by replacing them with a European ‘superdemos’. On the contrary, it must protect their diversity and their democratic self-determination. Thus, demoicracy is a genuine third way (Nicolaïdis, 2003, 2013) – or, according to the insightful metaphor used by Nicolaïdis (2013), a decision neither to remain on the northern bank of the Rubicon nor to cross it, but to navigate its waters instead. It is a third way in institutional terms, because demoicracy is a proper governing union (not an intergovernmental forum), but it is inescapably different from a traditional federal union. It is a third way in democratic terms, because it is a form of democratic government (hence, the EU can be the kind of body which can have a democratic deficit problem), but one where the democratic equality that must be achieved is that among the different demoi or peoples, rather than individuals, of the EU. Demoicrats therefore reject the assumption that each democracy needs its own demos: a specific form of democracy can be upheld by a plurality of (interdependent, yet sovereign) demoi. The alleged democratic deficit of the EU must hence be assessed according to this demoicratic standard.

Is demoicracy really a third way, however? In this paper, I critically assess the claim that it is and issue a mixed verdict. ² I argue that demoicracy does possess distinctive normative features. However, an institutional choice between intergovernmentalism and federalism must be made in order to realise those very features. When we try to settle the question of what institutional shape the normative ideal of demoicracy should take, we must specify what it means for different demoi to not dominate each other. And depending on which choices we make on that front, demoicracy will either ground a very specific (and specifically demanding, or rich) way of practicing intergovernmentalism, or an equally specific (and specifically cautious, or thin) form of federalism. It cannot, however, ground an institutional model which is genuinely alternative to both.

The argument is structured as follows. Section introduces some distinctions and makes some conceptual points. In particular, it (a) distinguishes between the problem of the EU’s democratic deficit and that of its institutional form and (b) reflects on the interplay between empirical and normative claims in both debates. Section 3 attempts to reconstruct the very ideal of demoicracy, by both offering an exegesis of the current literature and filling some of its gaps. Section 4 contends that, depending on how one interprets the republican ideal of non-domination, the reciprocal non-domination of interdependent peoples, which demoicracy aims to realise, will necessarily imply a choice for either a federalism or an intergovernmentalism of sorts. Demoicracy constitutes, therefore, a distinctive way of capturing the ideal of equality between different yet interdependent peoples, but cannot ground a fully distinctive institutional model of multi-level polity (i.e., one that is neither federal nor intergovernmental). Yet, it can offer guidance to steer either intergovernmental or federal instruments in specific directions, so as to better embody the idea of democratic equality between demoi. Section 5 concludes by making two partial concessions.

Two debates and the normative–empirical interplay within them

In order to understand what the demoicratic proposal is, and whether it succeeds in the aims it sets for itself, we must first have a clearer grasp of the debates to which it seeks to contribute. As mentioned in the introduction, two interdisciplinary debates dominate EU scholarship. The first concerns the federal or intergovernmental nature of the EU; the second its alleged democratic deficit. Doing full justice to these long, incredibly rich and multi-disciplinary debates here is simply impossible; I shall limit myself to sketching the kernel of the problems that sits at the centre of each of the two, respectively.

The question asked in the first debate is whether we should envisage the EU as a forum of sovereign states or as a federal union of sorts. Federalists advocate an EU some of whose rulings apply to EU citizens directly and trump domestic laws and regulations. In other words, the EU should have some sovereign powers of its own. In a federation, states relinquish full control over the possible developments and implications of their cooperation – and in so doing, they may turn out to be obliged to do things against their preferences as it is not in their power to stop decisions (Nugent, 2006). Federalism thus takes inter-state relations beyond cooperation into integration, and thereby involves some loss of national sovereignty. Diagnostic or empirical advocates of EU federalism typically argue that the EU already has some such features (and, often, that it should acquire even more); normative advocates that it should. The reasons why it should, in turn, may vary. Some may be purely strategic (e.g., enhancing the geopolitical power of the European region) or problem solving (e.g., establishing a fiscal union capable of delivering important social services in a context where this is increasingly difficult for states). Others may be based on the solidaristic aim of strengthening a European identity, thus overcoming nationalism. For the EU to be a federation, it needs to be able to take some decisions without the unanimous consensus of its member states, who therefore should not have veto power over all matters (i.e., they should relinquish full control). For those who argue that the EU already is a federation, this is already the case in some areas where qualified majority voting applies – and these are sufficiently ‘core’ areas to warrant a federal interpretation, although more integration of this kind is welcomed.

Intergovernmentalists contend, instead, that the EU should be conceived of as a ‘Europe of states’ whose governments cooperate to offset negative policy externalities as well as to enhance their internal problem-solving capacity (Moravcsik, 1993), whilst retaining full control over the conditions and consequences of their cooperation. Again, empirical arguments point out that this is what the EU is, as any accurate analysis of its institutions and laws shows it not to have any meaningful sovereign powers of its own. It is true that qualified majority voting applies in several areas, but as long as member states are free to leave, their not doing so can be taken as an expression of their overall preferences. Normative arguments for intergovernmentalism, in turn, are based either on its virtues (since international coordination can augment, rather than restrict, the ability of governments to achieve domestic goals (Moravcsik, 1993) or on the inappropriateness of federal conceptions for the European context. It is argued, for instance, that Europe lacks a demos with the shared public culture that is necessary to sustain a federal union. Alternatively, the claim is put forward that when the institutions that govern over us as so high up and remote from us, the risks of technocratisation and bureaucratisation are prohibitively high. Normative intergovernmental arguments can sometimes even acknowledge that some steps towards a more federal state of affairs have already been taken, and advocate their undoing (Streeck, 2014). Predictably, intergovernmentalists argue that consensus should be the norm for most or all decision making within the EU. Federalists often argue that intergovernmentalism fails by its own standards. Its main aim is to allow states to solve common problems and thereby amplify their internal problem-solving capacity; however, uncoerced consensus is often hardest to achieve precisely on those issues where a solution is most needed, as crucial interests often diverge. When this occurs, the risk is either to agree on highly suboptimal solutions or for multilateralism to degenerate into power politics.

The debate on the democratic deficit consists in a cluster of different complaints about the lack of democratic legitimacy of the EU. These are related to one another but only up to a point: there is, in other words, no single definition of what the democratic deficit is to begin with. ³ The first set of complaints points to the increase in executive power and a decrease in national parliamentary control in the EU area, whereby the increase of executive power occurs both at the member state and at the EU level, with the two being of course mutually reinforced. A key aspect of this phenomenon at the EU level is the power to initiate legislation which the European Commission – an unelected and arguably unaccountable body – possesses. A key aspect of this first set of complaints is that the EU drives a wedge between EU citizens and those who govern them – both in terms of perceived distance of the former from the latter and in terms of a ‘drift’ of implemented policies from actual democratic preferences. Other complaints concern the weakness of the only properly democratic institution of the EU, the European Parliament, because of its highly limited legislative competences and because of the absence of genuinely ‘European’ elections (EU elections may have spectacularly low turnouts; campaigns focus on national issues more than on EU ones; etc.). Some further complaints identify consensual decision making as the chief modus operandi of the EU, and lament its undemocratic nature – for a consensus among governments may well go against what the majority of EU citizens want. Others still are preoccupied with the ‘no demos thesis’: a jurisdiction without a demos cannot take democratic decisions worth the name. Only the presence of a demos – through the mechanisms of public deliberation, debate, and preference formation that can uniquely occur within it – can confer democratic legitimacy to processes of decision making. A further, crucial concern is the lack of spaces of political contestation within the EU (Follesdal and Hix, 2006): democracy occurs, not when policies simply ‘happen’ to match citizens’ preferences (and even that might not occur anyway; see the first set of complaints) but when they do so because they are responsive to them in a way that is allowed by genuine mechanisms of contestation. Finally, some complaints are more specifically concerned with the anomalous structure of the EU. It is argued, in particular, that both equality among EU member states (as embodied in the Council and in all consensual decisions taken by the EU) and equality among EU citizens (as embodied by the European Parliament and qualified majority voting) could, in principle, constitute a form of democratic legitimation – and that the democratic problem of the EU lies in the fact that it currently combines elements of both in contradictory and sometimes mutually defeating ways. ⁴ This is what is sometimes called the ‘structural’ democratic deficit of the EU.

Often enough, the two debates are considered to be one and the same thing. Advocates of ‘more’ democracy within the EU often argue in favour of a stronger role for the European Parliament as the voice of the European people and are therefore usually perceived as defenders of both democracy and supranational sovereignty or federalism. The case for federalism is taken to be a case for democracy writ large, capable of implementing decisions that are taken by, and binding for, all European citizens. Federalists often argue that transforming the EU into a proper federal polity would also settle the question of its democratic accountability – for EU citizens, like citizens of any federal union, would finally be able to exercise democratic control over the institutions that apply to them through the normal channels of any liberal democracy. Intergovernmentalists, instead, typically take two routes. Some contend that the EU has no democratic deficit problem to begin with (Moravcsik, 2002) – largely because (like any other intergovernmental forum) it is not an organisation apt for democratic accountability; it is the governments of its member states that are accountable to their citizens for the kind of foreign policy they pursue within it. Others argue instead that its democratic deficit lies precisely in its having problematically moved beyond its intergovernmental vocation (for instance, on matters of foreign or monetary policy), because, in so doing, it has unduly limited the room for democratic self-determination of its members states. The disagreement between federalist and intergovernmental conception of the EU seems therefore to be largely about the role and place of democracy within it. Thus, both camps often end up suggesting that their favoured solution to the institutional shape of the EU is also a solution to its democratic deficit. The ‘structural democratic deficit’ thesis seems to further strengthen the idea that the two debates are about the same issue. This thesis suggests that the lack of democratic legitimacy within the EU is due to the fact the EU is neither a supranational union proper nor a pure intergovernmental organisation. If it were the former, the democratic principle of equality among citizens would have to prevail; if it were the latter, the same would be true for the (intergovernmentally democratic) principle of equality of all states – and it would be the business of each and every state to respect the democratic equality of all its own citizens. By having a mix of both, the EU is trapped in a system whereby, because of the unclear and sometimes conflicting nature of the two principles, some actors and constituencies are systematically underrepresented and other systematically overrepresented (Bellamy and Kröger, 2013).

Finally, it is worth highlighting a striking feature of the literature on the institutional shape of the EU: we are not facing two completely parallel debates – one on whether the EU is and one on whether the EU should be federal or intergovernmental. Empirical scholars mainly attempt to offer a diagnosis of what kind of polity the EU is – but, in so doing, they usually also make claims about the desirability of certain institutional features and reforms. They reflect on what the current key institutional aspects of the EU are and attempt to reconstruct, interpretively, its historical trajectory. In so doing, however, they almost always provide assessments on which further institutional developments it would be wise and/or desirable to pursue (or not). Intergovernmentalists typically claim that it makes sense for the EU to be or remain intergovernmental ⁵ ; warn against not sufficiently thought through processes of further integration (Scharpf, 1999); and even pledge for the undoing of some of its steps (Streeck, 2014). A similar intermingling of empirical and normative arguments prevails among federalists. Historical arguments about the original mission of the EC/EU are used in order to push for a reinvigoration of the process of integration during sluggish phases; or the diagnostic claim is made that certain contradictions within the current structure of EU can only be solved if the already advanced process of integration is ‘completed’. On the other hand, normative political theorists working on the EU usually refrain from developing normative standards that are parachuted on the EU from a Platonic realm of ideals. On the contrary, they are almost always motivated by the idea that one should critically assess the EU by employing, as much as possible, the standards that the EU has set for itself. For instance, they typically point out that their normative recommendations rely, among other things, on an interpretation of the original mission of the process of European integration, of the will of the European people (or peoples), or both. ⁶ Thus, for instance, normative federalists often lament the slowness of the process of integration largely because, among other things, they think that it is a betrayal of the very mission of the EU. Nicolaïdis, in arguing for demoi-cracy, also embraces this intermediate standpoint, and calls it ‘normative inductivism’ (2013: 357): a normative-inductive approach is one which tells us how the EU ought to be on the basis, not of first principles alone, but also of ‘the deep texture of European history, law, and politics (Nicolaïdis, 2013)’. In a language more familiar to political theory, we might say that most normative theories of the EU are to some extent practice-dependent (Sangiovanni, 2008): they ground an account of what the EU should be on, among other things, an interpretation of its point and purpose from the point of view of its participants. As long as such point and purpose is not independently objectionable – as long as it is not, for instance, incompatible with independent and relatively uncontroversial moral principles, such as minimal ideas of equal concern and respect for all moral agents (Sangiovanni, 2008: 163) – it should be the guiding criterion to formulate normative and institutional recommendations. This seems to reflect the conviction that Europe belongs to Europeans – their history, their commonality, and what they fundamentally want to make out of the European project. If, say, we cannot reasonably detect a grain of federal spirit in the ‘deep texture of European history, law, and politics’, then our theories of EU justice, legitimacy or democracy should reflect that. This, of course, does not mean that the EU should be what the (majority of) Europeans want it to be at a given time, period. An institution can fail to be faithful to its point and purpose on grounds of, for instance, weakness of the will or a temporary prevailing of sheer selfish interests on the part of its members.

Demoicrats vindicate the demoicratic ideal in the same practice-dependent or normative-inductive spirit, i.e., as something which constitutes the original mission as well as the ongoing shared goal of the process of European integration – albeit one which Europeans have failed time and again. ⁷ Since the demoicratic literature shares this feature with the broader debate on the institutional nature of the EU, I shall also work within this paradigm. ⁸ My claim shall be that, even if demoicracy is the right practice-dependent ideal – i.e., even if it captures the ‘spirit’ of the EU correctly – realising it might necessitate a choice between an intergovernmental and a federal institutional form. As such, the argument of this paper might leave unmoved those who think that normative and institutional ideals should be completely independent of such interpretive considerations, and only mirror-independent moral principles.

The ideal of EU demoicracy

The previous section has established two points: (1) both intergovernmentalists and federalists claim that their favoured institutional proposal is also a solution to the democratic deficit of the EU; (2) both debates are characterised by the interplay between normative and empirical claims in a normative-inductive or practice-dependent spirit. The advocates of demoicracy follow the lead of the debate with respect to both (1) and (2), but make a crucial further move: they argue that both intergovernmentalists and federalists are wrong about what institutional form the EU should take, because they misinterpret the type of democratic ideal that the EU should aspire to embody. In other words, they fail in their interpretive exercise. The EU should neither be the expression of the equal democratic power of all European citizens nor a forum for multilateral cooperation which leaves all meaningful democratic questions at the domestic level. Instead, it should be conceived as the project to achieve a ‘Union of peoples who govern together, but not as one’ (Nicolaïdis, 2013: 351). And Nicolaïdis is adamant that this is not just a disembodied normative ideal, but a claim based on the rich history of the process of European integration. The creation of a ‘Union of peoples who govern together, but not as one’ is the most plausible reconstruction of the point and purpose of such an integration process.

What does this mean, however? As has been noted before (Lacey, 2015), whilst the demoicratic ideal as the appropriate model for the EU has been getting some traction as of late, the concept as well as its institutional implications remain largely vague – due in part to the fact that the term is not used in an identical way by all its advocates. ⁹ In this section, I provide an account of the demoicratic ideal which attempts to remain faithful to the motivations of its advocates – and indeed largely draws on their work – but offers a more unified and comprehensive vision of what it constitutes and aims to achieve. In so doing, the present section offers not only a description but also an active reconstruction of the concept of demoicracy, which partly draws on exegetic works, partly attempts to fill conceptual gaps in the current debate.

As the unconventional name suggests, a demoi-cracy is a proper union with governing features (a kratos), but its ‘citizens’ are the macro-agents constituted by the different peoples (or demoi) of Europe, rather than EU individual citizens taken as a unified European demos. A demoicracy, unlike an intergovernmental organisation, is a kratos proper, but one that is jointly held by the different democratic peoples of the EU. The demoi of Europe, under this interpretation, do more than merely pursue mutually beneficial multilateral strategies through their governments/states. They instead reciprocally acknowledge that they: (1) have mutual obligations to respect one another’s freedom qua demoi and that they (2) unavoidably affect one another (qua demoi) in problematic ways. They recognise, in particular, that their internal democratic processes might have problematic democratic externalities for each other: some decisions taken by a demos might undermine the capacity of other demoi to exercise their own internal democratic self-determination. ¹⁰ They thus acknowledge, not only their mutual interdependence broadly speaking, but also their democratic interdependence – namely, that they have the power to affect the health of one another’s democratic nature and that this generates reciprocal obligations. This initial definition already points out two important differences between demoicracy and traditional liberal intergovernmentalism. First, demoicracy is more than enlightened self-interest. It does not stem from the realisation that mutual benefits may come from cooperation, and that it therefore makes sense to cooperate (Moravcsik, 1993); it is instead grounded in the recognition of mutual responsibilities. A demoicratic EU is not a forum where different state actors come together to bargain, even if in a multilateral way. It is instead the place where more stringent and binding reciprocal commitments between the peoples of Europe ought to be entrenched (Nicolaïdis, 2013). Although the ground upon which the demoi of Europe have these mutual obligations is not always unpacked in the demoicratic literature, it is fairly clear that it is based on the idea that, when peoples are constituted as demoi, i.e., when they give themselves democratic institutions, then they are owed respect as collectives by other actors. When a people is also a demos, it is owed respect – and, in particular, its democratic decisions are. Again, although the point is hardly ever explicitly made, it is evident from discussions that this is due to the fact that, when a collective organises itself democratically, respecting it also means respecting the individuals which form it – hence respecting it does not mean making a choice between the freedom of collective agents and that of individuals. ¹¹ Second, whereas demoicracy shares with intergovernmentalism the premise that the EU should achieve the equality of certain collective agents, rather than of individual European citizens, it has a distinct account of which collective agents are at stake. Demoicracy advocates the equality of European peoples at the EU level, as distinct from both the equality of European citizens and the equality of member states. The aim of the EU is not merely to foster the problem-solving capacity of its member states, but the democratic capacity of its peoples. Therefore, the governments of member states should not be taken at face value if they do not suitably channel the voice of the people they represent. Thus, even if demoicrats believe that the EU should be respectful of the sovereignty between the polities which constitute it, this is compatible with the EU imposing demands on the quality of domestic democratic processes. Thus, both differences highlight the importance of the quality of domestic democracy for demoicrats.

In this spirit, EU demoi establish a union to guarantee their joint and reciprocal freedom qua demoi, i.e., their joint capacity to be demoi in a proper sense. It is crucial to point out at this juncture that this reciprocal commitment is cashed out by several demoicrats (Bellamy, 2013; Nicolaïdis, 2013) in distinctively republican terms: the kind of freedom which the EU demoi aim to obtain is not their reciprocal non-interference, but their reciprocal non-domination. ¹² Republicans define freedom as the condition of not being subject to the arbitrary or uncontrolled power of another agent – they object to what they call alien control, not to any control or interference at all. Domination occurs when the power between two agents is deeply unbalanced, to the effect that there are no effective constraints on its exercise: the dominating agent can act with impunity and without being accountable to the agent upon whom such power can be exercised (Pettit, 1993, 1997). Crucially, an agent dominates another if she is in such a position of asymmetrical power regardless of whether she decides to exercise it or not. Thus, interference can occur without domination, and domination can occur without interference. ¹³ EU peoples realise that their democratic interdependence is such that may enable some of them to dominate others in their democratic capacity. This, for demoicrats, reflects both the history of the European region –where some peoples have been able to subjugate others over and again – and its present set up, where high levels of interdependence put some democratic collectives in a position that enables them to exercise alien control over others. ¹⁴ Demoicratic institutions should enable the European demoi to regain control over the quality and health of their own domestic democratic institutions. For domination to be avoided, however, some non-arbitrary interference will be necessary. Non-domination is obtained when agents enjoy the guarantee that others cannot exercise alien control over them – and this typically occurs when a system of binding, justified and democratically authorised rules is in place. Thus, alien control of some demoi over others is avoided by establishing a form of joint and equal control over the conditions that allow their reciprocal non-domination. This resonates quite nicely with the idea that EU demoi are a Union of people who govern together, even if not as one. This is much more than intergovernmental multilateralism, because the demoi genuinely govern together on those matters – but on those matters only. In Nicolaïdis’s own words, demoi must enjoy ‘institutional and legal safeguards at the centre’ (2013: 363). However, if it is the reciprocal non-domination of demoi that must be guaranteed, the relevant system of rules and the kind of union that will follow from it must be faithful to what the union is an union of. If the union aims at protecting the reciprocal non-domination of the different (and inherently diverse) European demoi, then it cannot permissibly do so by replacing them with a European ‘superdemos’. On the contrary, it must protect their quality as different and diverse demoi, including, crucially, their entitlement to democratic self-determination.

If this occurs, demoicrats argue, demoicracy is a genuine third way to both intergovernmentalism and federalism. It is a third way in institutional terms, because demoicracy is a proper governing union (not an intergovernmental forum), but it is inescapably different from a traditional federal union. And it is a third way in democratic terms, because it is a form of democratic government, but one where the democratic equality that must be achieved is that among the different demoi, rather than individuals, of the EU. To understand this point more precisely, it might be helpful to point out that the advocates of demoicracy understand their ideal as jointly providing an answer both to the democratic deficit problem and to the question whether we should have an intergovernmental or a federal Europe. It is an answer to the first because it tells us that, whereas the EU is highly imperfect from the point of view of accountability, it is to the demoi of Europe, rather than to the European demos, that it should be made accountable to. It could well be that the EU currently fails even by this standard – indeed, several demoicrats think it does, although with different degrees of gravity depending on the author. ¹⁵ However, it is according to this standard that the alleged the democratic deficit of the EU must be assessed. It is an answer to the question of the institutional form of the EU because it constitutes a departure both from supranationalist traditions that perceive the need for and expect the emergence of a European demos and from intergovernmentalist understandings that postulate limited powers for European Union (EU) institutions based on delegation by the national demoi. (Cheneval et al., 2015: 2). The national demoi do not delegate, but govern, together – however, they do so on matters that concern demoi, rather than EU citizens directly.

At this point, one might legitimately ask what it actually means for the demoi to govern together, albeit not as one. Unsurprisingly, this is where the demoicratic literature strikes many readers as being both less precise and less cohesive. Institutional recommendations are both fairly sketchy and relatively diverse. However, at closer scrutiny, some important elements can be identified. First, it is argued (particularly by Bellamy, 2013; Bellamy and Weale, 2015) that the key institutions of EU demoicracy should be neither the Commission, nor the Council, nor the European Parliament, but rather domestic parliaments. Demoicrats think that this is not a paradox. If the EU is not a federation, it is not the Commission or the European Parliament that ought to be strengthened in order to tackle the democratic deficit. However, if it is not an intergovernmental forum either, we cannot take governments at face value (as we tend to do with the Council); we must look, instead, at what happens inside the member states of the EU. Demoicracy is democracy among demoi, not states taken at face value – and as such, it is a precondition for it that the demoi of each member state, via their legislatures, actually participate in EU law and policy making. ¹⁶ Domestic parliaments should be in the business on scrutinising, deliberating on, and wherever possible even initiating EU legislation. They should be actors proper of EU policy and law-making. This also clarifies why the standards of accountability for a demoicracy are higher than those of an international organisation: the EU is accountable to its demoi, not to its member states. It is not enough that states be represented equally in the EU; we additionally need instruments to ensure that it is the actually different demoi that are thereby represented. As Bellamy and Weale (2015) put it, ‘when governments make commitments to one another about their future behaviour, they simultaneously need to be responsible and accountable to their domestic populations in order to retain their political legitimacy’ (2015: 259). On the one hand, this makes individual states highly responsible for their own fate in a demoicracy, for they each must ensure that their own demos take an active role in this respect by being involved in EU policy and law-making. ¹⁷ On the other hand, however, this is not enough. First, the Union is also to be held accountable if and when it creates obstacles to this dynamic – as it does, for instance, if and when it is impatient towards careful and lengthy scrutiny of EU politics by different domestic publics. Second, each state has an obligation to acknowledge that other states also have obligations to ensure domestic deliberation. Governments must acknowledge that being fair to each other means allowing each of them to undergo high-quality internal democratic processes at each stage of the relevant EU-level decision making (Bellamy and Weale, 2015: 260). Thus, the strengthening of the role of domestic parliaments in EU affairs in the name of demoicracy imposes responsibilities on the demoi with regard to themselves, on the institutions of the EU, and on the demoi with regard to other demoi.

More detail on which principles ought to govern the institutional shape of a demoicracy is offered by Cheneval and Schimmelfennig (2013), who argue that demoicracy requires indeed supranational institutions, but such that guarantee the freedom and equality of all its peoples (not merely states). If we were to draw on a domestic analogy – and therefore consider demoi in a supranational settings to be relevantly like individuals in a national setting – we might conclude at this point that the freedom and equality of demoi is in principle compatible with a federal union. Only die-hard anarchists believe that citizens are unfree simply because they are subject to a sovereign legal and political order from whose authority they cannot escape. The consensus is that they are not unfree provided that the sovereign order in question has certain features – i.e., that it respects their fundamental rights and is under their democratic control. Why should this not be true for peoples, as well? This line of argument, however, seems to disregard the fundamental difference between citizens and demoi, and therefore to betray the very ideal of demoicracy. The freedom of an individual citizen must mirror her nature as an individual person, and this (anarchists excluded) is widely considered to be compatible with the presence of a coercive and non-optional political order, provided that it has certain features. Actually, most republicans would point out, in a Kantian spirit, that it requires such an order, for only political authority can establish the kind of regulatory framework and legal certainty that are necessary to avoid domination among individuals (Kant, 1999[1797]). At the level of demoi, one might argue that things are different. The default understanding of sovereignty – again in a Kantian spirit – is that a legitimate state may not be subject to coercion, because it is itself the authoritative source of coercive rules (Kant, 1999[1797]). In a democracy, a state is legitimate when it is democratic. This means that the demoi are the source of ultimate sovereign power. Respecting the freedom of a demos, therefore, means respecting the sovereign rule it establishes. This means that, for a supranational institution to be a demoicracy, it must respect all the demoi which constitute it qua ultimate sources of ‘pouvoir constituant’, i.e., the power to constitute a political order and to give themselves a constitution (Cheneval and Schimmelfennig, 2013: 342). If democratic peoples are not recognised as ultimate sources of pouvoir constituant by a supranational political order, they are disrespected qua peoples The supranational institutions of the EU therefore do not have pouvoir constituant in and of themselves, but rely on that of its member peoples. Therefore, a demoicracy can have supranational institutions (which do way more than facilitate multilateral negotiations), but only such whose basic conditions are ultimately under the control of the different demoi. Crucially, the demoi must have ultimate control on entry, exit, and some other basic rules: no demos may be obliged to join or stay in a demoicratic order by the decision of a branch of government only, by majority decision of a group of states, by majority decision with the participation of citizens that are not members of the democratic state in question. (Cheneval and Schimmelfennig, 2013: 342). Note how this resonates with Bellamy and Weale’s argument: it is the ultimate source of sovereign power within a democratic state, i.e., its legislature or parliament, which retains authority over demoicratic matters. Thus, we can conclude that neither intergovernmentalism nor federalism respect demoi qua demoi: the former unacceptably takes governments at face value, whilst the latter disrespects demoi qua sources of pouvoir constituant.

Finally, it is worth noting that, whereas I think this to be the most plausible reconstruction of the demoicratic ideal, some advocates of demoicracy think that it is a mistake to put too much emphasis on specific questions of institutional design. Nicolaïdis, for instance, suggests that demoicracy is an attempt to be faithful to the ideal of non-domination between the peoples of Europe rather than a specific institutional crystallisation. The same EU, with the same institutions, has been more or less capable of measuring up to the demoicratic ideal at different historical junctures. Demoicracy is an expression of mutual solidarity between the peoples of Europe who, however, simultaneously also acknowledge their inescapable differences as well as their potential to constitute a threat to one another. It is this solidarity cum difference that defines demoicracy from intergovernmentalism and federalism. As such, demoicracy is ‘normatively antithetic to both’ institutional forms even if it may ‘empirically borrow’ from either or both at a given time (Nicolaïdis, 2013: 353; emphasis in the original). ¹⁸

Can demoicracy be a third way?

The last section has put forward an interpretation of the demoicratic ideal. In light of this, is demoicracy, so construed, a genuine third way? In tackling this question, I shall grant a series of points to the demoicratic camp. I shall grant that normative and institutional ideals for the EU should be formulated in a normative-inductive or practice-dependent manner. I shall grant that different demoi can value their own independence and yet recognise that they produce democratic externalities, and that this can only be addressed by establishing a form of joint governing of sorts. Finally, I shall grant that the republican ideal of non-domination is the best way to cash out this aim. Admittedly, I find this easy to do as I agree with demoicrats on all of these points; however, my aim here is to isolate one conceptual question, namely whether demoicracy constitutes a distinctive ideal. This section argues that demoicracy does have distinctive normative features – but that, when it comes to pinning down how it should be institutionally realised, a choice between (rich) intergovernmentalism and (thin) federalism must be made.

The first building block of the argument is the following: demoicracy is, first and foremost, an answer to the democratic deficit problem, not to the issue of the institutional form of the EU – and the two, unlike what most participants in the debate assume, are not one and the same thing. The democratic deficit thesis argues that the institutions of the EU are not democratically legitimate – or not to a sufficient degree. Whereas the thesis is most vocally advocated by federalists in order to push for a strengthening of the European parliament and the transformation of the EU into a federal democracy, the issue is, as we have seen, more complex. A better system of democratic accountability within the EU can be achieved by creating forms of direct EU-level accountability to European citizens, where the latter are conceived of as members of a single European people (and demos). However, it can also be tackled by making sure that each and every existing national democratic public (and not just their governments) has the last word on every important piece of EU law-making – in other words, the aim can also be reached through less, rather than more, integration. Conversely, advocating federal integration need not entail advocating EU democratisation. Just to take the most obvious example, the European Commission and the European Parliament fare equally well from the point of view of their federal or supranational pedigree, but very differently by democratic standards.

The demoicratic ideal is an attempt to show that the current framing of the debate on democratic deficit is misguided, because it presupposes that democratic legitimation can only occur through the legitimation of binding norms by one single demos. Therefore, one should either create one such demos at the EU level or stick to a Europe of national democracies. Tertium not datur, because democracy requires a demos. The ideal of demoicracy, instead, tells us that it is possible for a plurality of demoi to govern together democratically over their reciprocal relationships, rights, and obligations. In other words, demoicracy gives us an alternative story about which agents ought to be in mutual relationships of freedom and democratic equality within the EU – demoi, not individual citizens. This means that it is possible to have a democracy without a single demos, namely a democracy regulating the joint democratic government of inescapably different yet also inescapably interdependent demoi. This is what demoi-cracy means: a democracy where the demos is a demos of demoi, not individual citizens. Yet, the idea that the demos is constituted by separate demoi is to be taken seriously: thus, demoicracy will necessarily be a form of democratic government which respects the right of each demos to its own internal process of democratic self-determination. To sum up:

Demoicrats care about the freedom and equality of the different demoi (not citizens) of Europe,

Demoicracy, thus characterised, has indeed distinctive normative features which contribute to an ideal of democracy and solidarity among the (democratically organised) peoples of Europe. It does not, however, constitutive a wholly distinctive ideal through and through because its most suitable institutional incarnation remains a largely open question – one which raises, in particular, issues concerning what it means for different demoi to be in reciprocal relations of non-domination. Depending on how we settle the question of what non-domination means, I suggest, the institutionalisation of demoicracy is either drawn towards very moralised and rich conceptions of intergovernmentalism, or towards meaningfully limited and thin forms of federalism – but not towards a third and fully distinct institutional construct. Thus, not only is demoicracy open to different institutional incarnations; what is more, these incarnations are variations of either intergovernmentalism or federalism, rather a genuine third alternative in institutional terms.

To understand this point, a brief reflection on what the non-domination of demoi might mean is necessary. Republican non-domination is typically conceived of as being best protected when both democratic control and institutionally sanctioned guarantees (for instance, in the forms of the rule of law and/or constitutional rights) are in place. The former ensures that the people exercise control over the kind of interference that is exercised over them (recall that non-domination is not the absence of interference), whereas the latter ensure that certain forms of interference that are incompatible with any understanding of non-domination, or certain forms of protection that are necessary for any understanding of non-domination, be in place at all times (say, freedom of expression or the right to strike). However, depending on the relevant brand of republicanism, different emphasis will be put on either side. Under a certain understanding of republicanism, democracy is the central value: no ‘guarantee’ can be such sub specie aeternitatis – only the people can determine whether this is the case or not (Bellamy, 2007). Solid forms of legal and constitutional protections which aim at determining what non-domination requires substantively are valid safeguards against non-domination only if they enjoy democratic legitimation and can be democratically revoked. Anything else, especially in the face of inescapable disagreement on these matters, means creating a de facto unaccountable class of ‘experts’ – be they policy makers, judges, or both. Thus, whereas a liberal society (i.e., one which subscribes to liberal rights) without democracy is in principle possible, republicanism can only occur if the conditions that set the terms of the joint and reciprocal non-domination of all are always under the control on whom they apply. This brand of republicanism delivers distinctive policy recommendations at the domestic level (such as the rejection of strong constitutional review and of constitutions which cannot be democratically revoked in a meaningful way) but is also more inclined to have a specific understanding of what non-domination between demoi might mean. Democracy plays the lion’s share in protecting against domination both at the inter-individual/domestic level and at the inter-demoi/international one. Under this democratically demanding reading of republicanism, a demoi is non-dominated when it can decide for itself, period. The demoicratic story is still true, in that European demoi might acknowledge that they affect each other’s democratic nature in problematic ways, and should recognise that they have an obligation to cooperate with the explicit purpose of avoiding that. However, the norms which this cooperation establishes have a republican pedigree if and only if the ultimate authorisation comes from each of the demoi on an ongoing basis. This brand of republicanism, in other words, puts forward a specific account of what is special about a democracy among different demoi. If such democracy is to respect the nature of its ‘citizens’ qua demoi, it must respect that they and only they are the ultimate source of ‘pouvoir constituant’. This means that the EU-level ‘demos of demoi’ cannot have supremacy over the single demoi. The point is to create an association which guards against ‘the domination of one people by another by preserving the capacity of the associated peoples for representative democracy’ (Bellamy, 2007: 499). This is a form of intergovernmentalism through and through. Bellamy (2013) and Bellamy and Weale (2015) unapologetically recognise that: demoicracy is thus equated to a form of republican intergovernmentalism. Republican intergovernmentalism ‘promotes […] the possibility for all individuals to live in representative states that possess democratic systems where collective decisions are made in ways that show them equal respect and concern through being under their public, equal control’ (Bellamy, 2013: 507). In other terms, republican intergovernmentalism is a particularly robust, and indeed moralised, form of practicing the intergovernmental game. It recommends that people associate to achieve certain moral goals and obligations, rather than win-win situations of mutual benefit. Under this understanding, demoicracy clearly is not an institutional third way – it is rather an exhortation to practice intergovernmentalism in a particular way.

The brand of republicanism which is usually identified as neo-roman, instead, puts a much stronger emphasis on the rule of law and on robust guarantees which are resilient to what (arguably volatile) majorities might want at a given time. Philip Pettit has developed a version of this brand of republicanism where democracy arguably plays a fairly modest – merely ‘editorial’, rather than ‘authorial’ – role (2004, 2013). In editorial politics, the emphasis is on the capacity of citizens to contest, but not necessarily make, the law. Authorial politics is the more creative realm of political activity – but is also the one where populism and irrational political instincts (including tyranny of the majority dynamics) prevail. Editorial politics, instead, is the realm of critical contestation of specific policies and laws – it is more reactive, but also more reflective. One does not need to go this far – and might be way more sympathetic to authorial democracy than Pettit – to point out that the more radically democratic variant of republicanism might be too cavalier with respect to legal and possibly constitutional guarantees. Domination occurs where I have no control over the rules that apply to me, but it also occurs when the democratic quality of my polity is generally high, and yet genuine democratic majorities take decisions that allow forms of what might be reasonably called ‘arbitrary’ interference. In other words, a strong concern for the risk of a ‘tyranny of the majority’ is and should remain a central republican concern. If my right to practice my religion is constantly threatened by democratic majorities, I am exposed to domination. And the same is true if it is simply exposed to democratic volatility. Thus, for neo-roman republicanism, the rule of law and some constitutional guarantees are central indeed (even if not necessarily with the same disregard for authorial politics which Pettit exhibits). Republican freedom requires democracy and robust unconditional guarantees of certain fundamental rights: both are necessary to ‘be one’s own master’. This has important and fairly self-explanatory implications at the domestic level, but it also have some equally central and not so obvious ones when it comes to non-domination among demoi. If demoi retain strong opt-out clauses on all important matters, there is a sense in which upholding conditions for the non-domination of the different demoi is up to the good will of all and each of them. This, however, might be seen as somewhat of a contradiction in terms. It is true that the non-domination of a demos must exhibit some form of respect for it being an independent polity, and that this is what we ultimately want to protect. However, we want to protect it because demoi can dominate each other. And if each demos is potentially capable of dominating others –and by definition it is, for this is what triggers the demand for a demoicracy to begin with – then the demoi should entrench some constraints on their capacity to do so. This cannot be up to their own good will: as we have seen, domination is matter of power, not intention, and non-domination consists in depriving certain actors of their capacity to exercise arbitrary power or alien control, not in them being benevolently inclined towards not using their power whilst still retaining it. If the demoi of Europe recognise that their capacity for reciprocal domination and for the production of mutual democratic externalities is a reality, and if they acknowledge that they have an obligation to address this, then addressing it means binding each other in meaningful ways. Obviously, these binding rules should not be something which an unaccountable EU bureaucracy decides and administers at will; they should, instead, be the product of a joint democratic government of the peoples of Europe, who decides on those rules without merging into a single demos. But such rules should be binding. This is what ‘institutional and legal safeguards at the centre’ (Nicolaïdis, 2013: 363) means. Nicolaïdis seems to acknowledge this point when she argues that ‘Germans and Greeks should not only have the right to put the problems they create for each other’s democratic health on each other’s political agenda, but should entrench institutional mechanism to address them’ (2013: 356; emphasis added). This, however, does require some form of federal union, even if one whose only area of competence is that of ensuring the necessary pre-conditions for the joint and mutual non-domination of EU demoi. A demand for entrenchment is a demand for the relevant demoi to bind themselves to a superior authority – even if that superior authority will only have very limited scope, and even if the whole point of this authority will be to guarantee the democratic health, diversity, and freedom of all demoi. It is a very thin form of federalism – but it is federalism nonetheless. It is worth noting that this form of binding is deeply in tension with some of Nicolaïdis’s other claims – and in particular, with her suggestion that, regardless of the specific institutions in place, the EU can fare better or worse in demoicratic terms depending on what it chooses to do with at some critical junctures. This suggestion that EU institutions are what its members make of them seems to involve a level of discretion that is incompatible with non-domination understood in these terms. Non-dominating each other, under this reading, seems again to be a matter of good will.

I therefore conclude that demoicracy is a distinctive and arguably appealing normative ideal – but that, institutionally speaking, demoicrats must still choose between intergovernmental and institutional variants. The specific understanding of republican non-domination that one endorses has implications for the kind of institutional structure that the demoicratic ideal should take. The radically democratic understanding of republicanism must be committed to the view that, for demoicracy to be distinctive, it must consider demoi to be the ultimate sources of sovereign authority – and this means that only form of joint governing between independent demoi can be intergovernmental in kind. For those who believe, instead, the non-domination requires robust guarantees, the demoicratic level of governance must be able to enforce at least the most fundamental terms of mutual non-domination on the member demoi. Under this understanding, demoicracy is still different from democracy in a single demos, but this difference is largely (although not exclusively) quantitative: it must respect that it is democracy over independent demoi, and therefore it must be suitably thin. Within those narrow boundaries, however, it will enjoy the supremacy of a federal union. However, this choice also means that, at closer scrutiny, the bifurcation between intergovernmentalism and federalism is inescapable after all.

Before I conclude this section, I would like to address one possible demoicratic rejoinder. Arguably, the variant of demoicracy put forward by Cheneval and Schimmelfennig (2013) is the closest to an institutional ideal that is neither recognisably intergovernmental nor recognisably federal. In their view, demoicracy has supranational institutions proper, but the demoi retain ultimate control on entry, exit, and some other basic rules. Supranational institutions do govern as long as the demoi stay in the Union, but demoi retain, for instance, a right to leave the Union at any point. Whereas it is true that this variant constitutes the strongest challenge to my claim, I think that it ultimately fails. If (and it is a big if) the right to exit is sufficiently strong and robust, I fail to see how this does not collapse into intergovernmentalism, although a particularly thick one – one according to which, as long as one stay in the association, one accepts a thick web of supranational rules, although the possibility to no longer accept them remains meaningful at any time. The ongoing justification for supranational institutions is intergovernmental – each and every demoi has the last word on their participation in them. This is incompatible with the second understanding of non-domination I have described. It is not compatible with the first, but the first brand also explicitly subscribes to demoicracy qua republican intergovernmentalism. Again, one might object that claiming that demoi can bind themselves in a way that allows for no return is a self-contradiction, for a demos who does that ceases to be a demos qua source of ‘pouvoir constituant’. I would argue, however, that this is an objection to the very idea that the second understanding of non-domination is applicable to demoi, not to my claim that demoicracy must decide between intergovernmentalism and federalism. If the objection were correct, republican intergovernmentalism would be the way to go, for the only form of non-domination which can possibly apply to demoi would be the radically democratic one. If that were the case, my claim that demoicracy is not a distinctive institutional ideal would still stand. I also think, however, that there is hope for thin demoicratic federalists to address that objection. Doing it lies fully beyond the scope of this paper, but it is clear that the objection is based on a very specific understanding of sovereignty –namely, as an all or nothing deal, as something which you either have or do not have. The possibility to divide different sovereign competences is, however, the subject of many complex and long-standing debates in law, political science, international relations, and philosophy. The jury is clearly still out on whether it is possible to relinquish some sovereignty without relinquishing it all. If it is possible, then it is possible for the demoi of Europe, as well. And whereas demoicracy, under this reading, would ground an importantly thin form of federalism – a very different one from the model which traditional European federalists advocate – it would still be a recognisably federalist union.

The point of this paper is therefore vindicated: whereas demoicracy has distinctive normative features, it does not constitutive a wholly distinctive institutional form: the demoicratic ideal must be translated either into a specifically moralised way of practicing the intergovernmental way or into a particularly thin federal union.

Conclusion

The previous section has argued that demoicrats must choose between intergovernmentalism and federalism – even if between particularly rich and moralised versions of the former or particularly thin versions of the latter. I would like to end this paper by making two partial concessions. I have argued that, if the necessary conditions for the joint and reciprocal non-domination of EU demoi must be robustly guaranteed through ‘safeguards at the centre’ (and we have seen that it is a big if even within the republican camp), there must be a sense in which the EU is the ultimate source of authority over those conditions – and therefore a sense in which it must be a federal union of sorts, if a very thin one. Some of its rulings must be binding and ultimate, and claiming that demoi must retain a strong right to exit at all times is in stark tension with this requirement. If (and again it is a big if) non-domination is understood in this way, when a demos acknowledges that it might dominate other demoi and enters the EU as a result, it binds itself in a way that is incompatible with a strong right of exit. ²⁰ Hence, the demoi relinquish full control over the conditions of their cooperation.

However, if the point of the EU is the equal respect of all demoi, two important qualifications must be made. First, it does indeed remain an open question whether the conditions that guarantee the joint non-domination of all demoi must be decided upon by consensus or by majority decision making – and, if the latter, majority among whom? As far as this paper is concerned, the possibility that unanimity must be required on certain or even many issues remains entirely an option – and if it turns out to be the right option, the EU will differ from other federal unions in a very crucial matter of decision making, and not just in terms of its thinness. What I have argued is that, if non-domination is interpreted in this specific way, a commitment to a Union of peoples does have an element of no return and unconditionality to it; but this leaves open the question of how the union should work if it is to be a union of demoi, rather than individuals. It could be that a federal demoicracy requires consensus on a much higher number of crucial matters than a traditional federal union. Settling this issue lies entirely beyond the scope of this paper, but it is clear that the combination of no (easy) right to exit and consensual decision making on many issues would be rather unusual. ²¹ If that were to be the right way to interpret and institutionalise demoicracy, there would be a thin sense in which demoicracy is a third way.

Secondly, if a demos is a source of pouvoir constituant, it is quite plausible that the acknowledgment of a duty to form a union of non-dominating demoi may only permissibly come from each demos itself. Hence, the acknowledgment is an acknowledgment of a duty of no return, but no demos can be forced to acknowledge this duty and enter the Union. Hence, even if we think that a EU demoicracy should be a federal union of sorts, the process of its formation must be more like that of Switzerland than like that of the USA. ²² Saying that the recognition of a duty to establish joint non-domination is a duty to form a federal union (with an element of no return) does not necessarily mean that demoi can be forced into it.

These two concessions (and the first in particular) do point to two senses in which demoicracy might be different from more familiar federal unions. However, leaving those aside, the argument of the paper still stands: demoicracy is a distinctive ideal of democratic equality and solidarity beyond borders, but it does not constitute an alternative between intergovernmentalism and federalism in institutional terms. This might well be what Nicolaïdis means when she claims that demoicracy is a model of inter-demoi solidarity rather than a specific institutional crystallisation. However, from a practice-dependent perspective, it is not entirely consistent to put forward a normative ideal without clarifying at least its most fundamental institutional implications (such as whether member states of a demoicracy will enjoy absolute sovereignty or not). If we do not know what such an ideal commits us to, it is not clear that we can fully grasp what it is, and whether we should subscribe to it or not (Banai et al., 2011: 54–55). ²³ Therefore, granting that demoicracy is only a distinctive normative ideal cannot leave demoicrats entirely satisfied.

Using Nicolaïdis’s metaphor, navigating the Rubicon is not an option: we can either stretch our arms as much as possible towards the other bank, whilst sill avoiding jumping into the water; or we can perform a very minimal crossing, which involves simply reaching the other bank and staying put rather than venturing into the valleys in front of us. But choose between one of the two banks we must.