Farewell to constituent power? The Conference on the Future of Europe, citizens’ assemblies and the democratic minimum

Abstract

In this article, I ask about the extent to which the Conference on the Future of Europe (CoFoE) has advanced democracy in the European Union. I critically engage with the claim that the CoFoE’s success should not be measured by whether it enabled constituent power, or ultimately results in treaty reforms, but by the fact that, by introducing citizens’ assemblies to EU politics, it has laid the foundation for participatory democracy in the European Union. Drawing on established theories of participatory democracy, I argue that this interpretation misses the point. To put forward an alternative view, I revisit James Bohman’s concept of a democratic minimum. The best democratic defence of permanent EU citizens’ assemblies is that they could provide citizens with the capacity to initiate deliberation about common concerns – and thus function as a nucleus for constituent power in the European Union. Nevertheless, the idea should be viewed with caution, as permanent citizens’ assemblies could just as well become a democratic fig-leaf allowing EU institutions to reject calls for fundamental reforms. Much therefore depends on their institutional design.

I. Introduction

Has the Conference on the Future of Europe (CoFoE) advanced democracy in the European Union, and if so, in what ways? The European Union has long been suffering from an oversupply of established institutions determining the course of European integration. Apart from the occasional ratification referendum, the European Union’s constitutional development, in particular, largely evades popular control. There is no way for citizens to initiate fundamental institutional changes. No restructuring of the European Union is possible without the consent of the member state governments, who – while often affirmed as the ‘Masters of the Treaties’ (Grimm 2017: 46) – in this context need to be treated as actors with institutional self-interests whose claim to legitimate authority is questionable. While the treaty-amendment procedures of Article 48 TEU provide for influential positions for the European Parliament (EP), the Commission, and the European Council (i.e. the national executives), nowhere do we find electoral mechanisms or forms of representation that would allow citizens to provide direct input on the design of the EU polity. Thus, at least since the European Convention of 2002/03, the question of how the process of European integration could be opened up to constitutional initiatives from below has been lingering above the European Union.

Against this background, the CoFoE presented itself as an extraordinary opportunity. When the Conference was first announced, it raised expectations of treaty change and institutional reforms. Together with the democratic rhetoric that accompanied the initial stages of the process, this led to the hope that the CoFoE could provide an opening for citizens to reclaim constituent power at the EU level. These hopes were quickly disappointed, however, when it became clear that control over the final outcome remained firmly in the hands of the European Union’s constituted powers, and when a number of member states declared early on that they would oppose any proposal for treaty amendments. At the same time, the CoFoE’s institutional setup lacked a clear conception of who ‘we, the people’ are in the EU context. Accordingly, the whole process took on a somewhat diffuse character and never developed the kind of momentum that would have been necessary to result in an exercise of constituent power. Nevertheless, the CoFoE is now receiving a lot of praise, suggesting that it may have been a democratic leap forward after all. A number of scholars have started to argue that the CoFoE’s success should not be measured by whether it enabled constituent power, or whether it will ultimately result in treaty reforms (which, by now, seems unlikely), but by the fact that it introduced a new form of (deliberative) participation in the European Union, which might be permanently institutionalized: the citizens’ assembly. The discussion has thus shifted from the merits of the CoFoE to the democratic potential of permanent citizens’ assemblies. The establishment of such institutions, we are told, would constitute a significant step towards EU participatory democracy. In this article, I criticize this view and, building on James Bohman’s (2007) idea of a democratic minimum, put forward an alternative interpretation of the democratic potential of permanent citizens’ assemblies, arguing that we should not give up on constituent power too easily.

The article is structured as follows. I begin by tracing the hopes for a return to democratic EU constitution-making that the CoFoE raised in its early phase. These were not simply brought to the European Union by external actors, but in fact prompted by the main initiators of the Conference (Part II). In the next step, I assess the CoFoE’s record in terms of constituent power. The picture is quite sobering. Neither did the European Union’s constituted powers transfer any decision-making power to citizens, nor did the procedures express a clear idea of who was authorizing whom to shape the future of the European Union (Part III). Then I turn to the view that in fact the CoFoE was not about treaty change, or constituent power, but should be seen as the beginning of EU participatory democracy, which would centre on permanent citizens’ assemblies. I argue that this interpretation exaggerates the transformative effect that such institutions would have on the ordinary workings of the European Union (Part IV). Building on Bohman (2007), I argue that the democratic potential of permanent citizens’ assemblies rather lies in the institutionalization of the democratic minimum. Providing citizens with the capacity to initiate deliberation about the EU polity, such forums could become a nucleus for constituent power in the European Union (Part V). However, the experience of the CoFoE suggests that they could just as well turn into a democratic fig leaf allowing constituted powers to reject calls for fundamental reforms. In light of this, I formulate lessons for the design of EU citizens’ assemblies (Part VI).

II. Initial expectations regarding the CoFoE: Return to EU constitution-making

The CoFoE was a one-year process of citizen participation in the European Union that started in April 2021 with the declared mission of a ‘bottom-up exercise for Europeans to have their say on what they expect from the European Union’ (European Union 2021: 2). The Conference concluded on 9 May 2022 with the delivery of its final report, including 49 proposals for EU reforms and new EU policies to the heads of the EP, Commission, and Council. Institutionally, the CoFoE consisted of three core elements:

1. 1. Multilingual digital platform. The CoFoE provided an online forum, which allowed individual citizens to submit ideas, which were supposed to feed into the deliberation in the more formal settings.

2. 2. European Citizens’ Panels. The CoFoE encompassed four transnational citizens’ assemblies, each composed of 200 randomly selected EU citizens from the 27 member states. These forums were structured as deliberative mini-publics, each focusing on a different thematic area.¹ In addition, citizens’ assemblies and other events were held at national and regional levels, which played a less central role.

3. 3. Conference Plenary. The Conference Plenary, composed of representatives of various EU and national political institutions as well as delegates from the Citizens’ Panels, was the culmination of the CoFoE. The Plenary’s task was to translate the inputs from the Citizens’ Panels and the digital platform into final proposals.

The CoFoE was not explicitly intended as a constitution-making process, nor was it officially meant to trigger one. However, from the start – from the first public ponderings of the idea – it was linked to hopes for fundamental reforms of the European Union. Indeed, these expectations were raised by the main proponents of the Conference, who signalled that for them this was about more than new directions in EU policy. In a 2019 op-ed, often seen as the initial ignition of the CoFoE (cf. Alemanno 2020: 488), Emmanuel Macron called for ‘a Conference for Europe in order to propose all the changes our political project needs, which is open even to amending the EU treaties’ (Macron 2019). In line with this, early in the process, Ursula von der Leyen and the Commission were committed to put institutional matters such as the Spitzenkandidaten system and transnational lists for European elections on the agenda (European Commission 2020: 2). The EP stated that a genuine follow-up on the results of the Conference should include treaty change if necessary (European Parliament 2020: 6). Fuelled by such statements and providing ‘a focal point of hope and anxiety for those with very different views as to the appropriate polity ambitions and the optimal political form of the EU’ (Walker 2023: 13), the Conference opened up the prospect of EU constitutional change.

Inevitably, this stimulated the political imagination of both civil society actors and academic observers. The Conference was initiated in a situation in which the European Union had already been confronted for several years with increasing attempts by citizens to reclaim constituent power (Patberg 2020: Ch. 2). Especially the Eurozone crisis, and in the build-up to the CoFoE also the COVID-19 crisis, have led to the view – persisting to this day – that the European Union is confronted with challenges that require reforms also at a constitutional level (Kreuder-Sonnen 2023; White 2023). Accordingly, Federico Fabbrini (2020: 413) argued, at a time when the exact form of the Conference was still in debate, that ‘if the whole process is to be more than simply a talking shop, it must necessarily lead to some structural reforms of the EU constitutional settlement’. Similarly, Andrew Duff (2019: 4), a previous Member of the European Parliament, declared that the central purpose of the Conference should be ‘to prepare the ground for a new Convention that will eventually be summoned to revise the EU treaties’. Citizens Take Over Europe (2021), a coalition of over 60 civil society organizations, claimed that the ‘Conference should have the freedom to set its own agenda and to put forward every possible proposal for the future of Europe, with all options on the table, including treaty changes, and without pre-emption of any of the outcomes of the discussion’.

Given these ambitions, the CoFoE was interpreted by some as indicative of a (re)turn to a democratic mode of EU constitution-making – and sparked new constitutional visions. As Paul Blokker (2021: 330) put it, ‘[t]he Conference on the Future of Europe potentially indicates a (partial) return to a constituent dimension in European politics’ (see also Ballangé 2022: 670). While he remained sceptical that the CoFoE could facilitate a full-fledged articulation of constituent power, he credited it with a principled openness to ‘reformists constituent articulations’ (Blokker 2021: 332). The CoFoE appeared as a way for citizens to impact the future of European integration in ways unavailable in the European Union’s daily business and impossible according to the ordinary rules of the game. Fabbrini (2020: 411–13) pondered what the outcome of this constituent process should ideally be and claimed that the CoFoE should aim at a political compact – a new intergovernmental treaty outside EU law that would allow willing member states to deepen European integration in areas where the unanimity requirement of Article 48 TEU currently prevents changes. In short, the CoFoE process was seen as a potential ‘constitutional moment’, a democratic opening that might lead – possibly in unconventional ways, and contrary to established procedures – to fundamental changes in the EU constitutional order.² Given these initial hopes, which were not just brought to the European Union from the outside, but in part nurtured by the main institutional supporters of the Conference, it is fair to ask: How did the CoFoE fare in terms of enabling constituent power in the European Union?

III. The reality of the CoFoE: A process controlled by constituted powers

Constituent power denotes the democratic entitlement and capacity of the people to give themselves a constitution and thus to determine the structure and competences of public authorities. To fully grasp its meaning, the concept needs to be considered in its opposition to the counter-concept of constituted power. The classical distinction between constituent and constituted powers – first systematically set out by Emmanuel Sieyès (2003: 136) – draws a line between the power to make a constitution and the power vested in the political institutions established by this constitution. Two important ideas are linked to this distinction. First, only the people as constituent power have the right to establish public authorities. In other words, constitutions should have a democratic origin. Second, constituted powers – the government, legislature, courts and so on – must not decide on their own terms of operation, which means that they should not be involved in constitutional politics (see Patberg 2017: 51–2). The underlying concern is to secure citizens’ political autonomy – that is, their capacity to govern themselves through the medium of positive law.³ While in the daily business of democratic states, the primary context of political autonomy is ordinary law-making, political autonomy remains incomplete unless it extends to constitutional law-making (Kalyvas 2008: 297–300). Citizens must be able to determine not only what laws they want to live under, but also what kinds of public authorities they want to be subject to, what procedures should be used to make collectively binding decisions and so on. Constituent power, then, describes political autonomy at the level of constitutional politics. As such, it is not limited to original constitution-making, or revolutionary forms of manifestation, but is relevant whenever the ‘meta norms’ that determine the structure and competences of a political system – including that of the European Union – are at stake.

The reality of the European Union, however, is that constituted powers operate as de facto constituent powers. This applies first and foremost to the governments of the member states. They are key players in the EU political system, yet also the main protagonists of EU constitutional politics. When they transfer competences to the European Union, they are putting them (partly) in their own hands. When they shape EU institutions, they are also deciding about their own role within them. At the same time, no constitutional change is possible without their consent. This is what I call the usurpation of constituent power in the European Union.⁴ The remedy would be a clear separation of constituent and constituted powers within the EU political system, so that citizens can take control of European integration and pursue EU reforms even against the will of constituted powers – for example, by means of a permanent constitutional assembly. A key puzzle, of course, is who exactly should be empowered here. Who are ‘we, the people’ in the EU context? In this regard, one can distinguish three competing models (see Patberg 2020: Part II): first, the regional-cosmopolitan model, which claims that constituent power lies with the political community of EU citizens; second, the demoi-cratic model, which claims that constituent power lies with the peoples of the member states; and third, the pouvoir constituant mixte model, which claims that constituent power lies simultaneously with EU citizens and the national peoples (or, in another variant of the model, the national citizens). Depending on which view one adopts, different requirements for a democratic process of reforming the European Union follow (e.g. who should elect representatives, for whom should they speak, how they should decide).

If we look at the CoFoE’s institutional setup based on these theoretical considerations, we encounter a structure that is still very much in conflict with the idea of constituent power.⁵ First, in terms of where the final power to decide was allocated, the procedures are a clear expression of the supremacy of constituted powers. Not only was the Conference placed under the authority of Commission, EP, and Council, acting as a joint presidency, but the Executive Board, also composed of representatives from these three institutions, was responsible for drawing the conclusions of the Conference Plenary, which in effect also means of the Citizens’ Panels and the online crowdsourcing. This turns the relationship between constituent and constituted powers on its head. The EU institutions apparently do not see themselves as trustees of revocable powers, dependent on the authorization of their citizens, but as legitimately taking the lead in shaping their own future. Here it is telling how Ursula von der Leyen (2019: 19) promised, in a generous tone, in her agenda for the European Commission 2019–2024, that the Conference would ‘bring together citizens … and European institutions as equal partners’ (emphasis added). And even that was promising a lot, because immediately after the signing of the Joint Declaration that established the Conference, several member states insisted that the CoFoE should not create any legal obligations for the European Union to implement reforms. Moreover, various member states were not prepared from the outset to accept treaty change of any kind. In short, while the European Union’s constituted powers consulted citizens on the future of Europe, they did so only on their own terms, without awarding them any decisional power.

Moreover, the CoFoE’s institutional setup was not based on any discernible idea of who ‘we, the people’ are in the EU context. Each of the three models mentioned above (regional cosmopolitanism, demoi-cracy, pouvoir constituant mixte) implies different ideas of how the CoFoE should have been organized – for example, whether the process should have been driven by representatives of European citizens, of national citizens, or both. However, the CoFoE remained completely ambiguous in this regard. The composition of the Conference Plenary was characterized by a broad but unsystematic inclusion of different actors, all of whom were allowed to send a certain number of representatives: the EP and the national parliaments, the Commission, the Council, the European Economic and Social Committee, the Committee of the Regions, the European Citizens’ Panels, the national Citizens’ Panels and Conference events, regional and local authorities, social partners, civil society organizations and the President of the Youth Forum. While the European Union presented this an expression of openness and inclusiveness, it leaves one wondering who was authorizing whom to shape the future of the European Union. In whose name was the Conference supposed to speak, and why should we assume that it was legitimized to do so? There never was a clear conception of who could claim a democratic mandate, and for what, who could be held responsible for results, and how. The EU institutions organized a process that appeared participatory on the surface, yet remained firmly under their control. The eclectic structure of democratic inclusion contributed to preventing any dynamic that could have resulted in challenges to the constituted powers’ dominant position. In terms of constituent power, then, the CoFoE’s record is quite sobering.⁶

IV. From the CoFoE to permanent citizens’ assemblies: The rise of participatory democracy in the European Union?

When it became clear that ambitions of democratic constitution-making would not be realized, a new perspective on the CoFoE emerged. Even if the CoFoE did not enable constituent power, it may have had other benefits. More precisely, the discussion shifted from the assessment of the CoFoE to the question of which institutional innovations it might indirectly pave the way for – and what their democratic potential is. Some scholars favourable to the project, most vocally Alberto Alemanno (2021), started to argue that ‘it’s not about treaty change, it’s about European democracy’. According to this view, constituent power is the wrong framework of evaluation for the CoFoE:

Rather than measuring the Conference’s success by its ability to gingerly lead to Treaty reform – or transmute into a new mode of constitution-making – it might be more relevant to gauge it against its capacity to offer both EU and national institutions as well as citizens a foretaste of a more intelligible, deliberative, and therefore more citizen-centric transnational Union. (Alemanno 2020: 25; see also Alemanno and Nicolaïdis 2021)

This new take meets with a theoretical literature that argues that there is no need for constituent power in the European Union in the first place, and that we should rather aim for more fluid forms of bottom-up democratization: ‘Rather than seeing the absence of a transnational constituent power as a cause of the European Union’s “democratic deficit”,’ we should regard it ‘as an opportunity for unfettered democratic participation’, we are told (Ballangé 2022: 655; see also Seubert 2021).

Specifically, the new interpretation of the CoFoE that we are being offered is that what really matters is the success of the Citizens’ Panels, which are presented as ‘the most innovative component of the Conference’s participatory infrastructure’ (Alemanno 2020: 506). The explicit hope is that the CoFoE could develop prefigurative power such that this format of including citizens in policy-making becomes a permanent part of the EU political system. Indeed, a proposal to regularly hold citizens’ assemblies made it into the final report of the Conference (CoFoE 2022: 79). Quickly, a number of scholars, in cooperation with the Bertelsmann Stiftung, presented a model for institutionalizing citizens’ assemblies as a form of ‘next level citizen participation in the EU’ (Abels et al. 2022). Likewise, the EP’s Committee on Constitutional Affairs (AFCO) commissioned a study outlining how the European Union could use citizens’ assemblies to meet ‘mounting citizens’ expectations for greater participation in EU decision-making’ (Alemanno 2022: 1). According to this model, which takes inspiration from the CoFoE, there would be two types of EU citizens’ assemblies: a permanent Citizens’ Chamber and temporary Citizens’ Panels, both composed of randomly selected EU citizens. The task of the Citizens’ Chamber would be to deliberate for which topics Citizens’ Panels should be set up (climate change, electoral reforms etc.), which would then work on concrete ideas for new EU policies. This process could be activated in both a bottom-up and top-down way – that is, initiatives could be brought to the Citizens’ Chamber by ordinary citizens, for example through petitions, and by the main EU institutions. The decision on whether to set up citizens’ assemblies and how to implement their recommendations would be left to the Commission, EP, and Council (Alemanno 2022: 63–65). Similar ideas for the design of a ‘European Citizens Assembly’ (ECA) have been advanced by Citizens Take Over Europe (2022). Roughly summarized, the ‘draft blueprint’ for an ECA puts forward two models: first, ad hoc citizens’ assemblies initiated from outside the EU institutions, for example by a permanent, randomly selected Citizens Board or through a European Assembly Initiative where citizens collect signatures; second, a permanent citizens’ assembly, which either operates largely autonomously or is connected to the wider public ‘through channels for participation on an online platform and social media’ (2022: 4).

The key notion in these proposals is participation. Supposedly, the CoFoE – more precisely, the debate about permanent citizens’ assemblies it has triggered – marks (an opportunity for) the rise of EU participatory democracy. The narrative supporting this view goes as follows (Alemanno 2020: 8–9). In 2009, the Lisbon Treaty made participatory democracy a fundamental principle of the European Union – in addition to representative democracy. Article 10 (3) TEU states that ‘[e]very citizen shall have the right to participate in the democratic life of the Union. Decisions shall be taken as openly and as closely as possible to the citizen.’ However, attempts to fill this clause with life, such as the European Citizens’ Initiative, have largely failed. ‘Most of the EU citizens’ participatory channels … remain unknown, scattered and underused by the average European citizen’ (Alemanno 2020: 9). Establishing permanent citizens’ assemblies is now presented as a way to finally provide the critical infrastructure for EU participatory democracy.

If we want to assess this view, the first question is what is meant here by participatory democracy. First, it seems clear that this interpretation of permanent EU citizens’ assemblies cannot be squared with the canonical contributions by Carole Pateman (1970) and Benjamin Barber (1984), for whom the primary purpose of political participation is an educative one: to develop skills and dispositions on the part of citizens that allow them to govern themselves. According to Pateman and Barber, this could only be achieved through a comprehensive societal transformation, including elements such as workplace democracy. What Alemanno and others have in mind when they refer to participatory democracy is – more down to earth: procedural reforms that award a more central role to citizens in decision-making processes. Thus, let me consider a more recent model of participatory democracy that goes in this direction and indeed envisages a key role for mini-publics, namely Hélène Landemore’s (2020) open democracy. Landemore’s model can be ‘considered a variety’ of participatory democracy, yet one that ‘is not premised on mass participation at all times’ (2020: 14) – and one that is anchored in the paradigm of deliberative democracy.⁷ According to Landemore, representative democracy has one major ‘design flaw’, namely ‘that it is centered on the principle of periodic elections’ (Landemore 2021: 69). In her view, this has a number of negative consequences, including the predominance of a partisan logic that runs counter to unbiased deliberation and a large influence of social and economic power on decision-making (corporate lobbyism, etc.), which undermines political equality. Above all, however, electoral mechanisms privilege the idea of citizens’ consent to rule over that of citizens’ exercise of political power (self-rule). Representative democracy, Landemore argues, creates a gulf between elites and citizens, yet pretends that the decisions of the former can be equated with the choices of the latter. Open democracy, by contrast, seeks to put ‘ordinary citizens, rather than elected elites, at the center of political institutions’ (Landemore 2020: 218).

The institutional key to this is meant to be ‘the “open mini-public”: a large, all-purpose, randomly selected assembly …gathered for an extended period of time … for the purpose of agenda-setting and law-making of some kind, and connected via crowdsourcing platforms and deliberative forums (including other mini-publics) to the larger population’ (Landemore 2020: 13). To prevent the negative effects of elections, Landemore aims at a combination of lottocratic, self-selected, and liquid representation to choose citizens for participation in the different arenas. She imagines ‘a constant rotating of ordinary citizens in and out of the variously nested and networked decision-making loci’ (Landemore 2021: 76). At the top of open democracy’s institutional architecture is a sovereign mini-public, empowered to make law, which is connected to various satellite mini-publics whose (non-binding) conclusions would be fed into the deliberations of the sovereign mini-public, as would the results of online crowdsourcing (2021: 83). Landemore does not take a definitive stance on whether these institutions should complement or completely replace the traditional setup of representative democracy, yet she invites us to ‘consider the scenario of an open democracy where elected functions are minimized or even absent altogether’ (Landemore 2020: 145, emphasis added). Following this model, a participatory democracy would significantly reduce the importance of some of the key practices of representative democracy, including voting, partisanship, campaigning, and parliamentarism.

If one looks at the idea of permanent EU citizens’ assemblies through this lens, one might indeed see such institutions as a first – although small – step in the direction of participatory democracy. The proposed assemblies do not follow an electoral or partisan logic, but are meant to give ordinary, randomly selected citizens a central place in deliberation about EU politics. Through the tool of online crowdsourcing, they could be made open to input from all EU citizens, not just those directly participating. In this way, permanent EU citizens’ assemblies would respond to some of the key tenets of open democracy. That being said, we must not over-estimate the actual expansion of opportunities for self-government that could be achieved. While Landemore seeks to include as many citizens as possible in deliberation, through a proliferation of mini-publics at all levels of the polity, the proposals for permanent EU citizens’ assemblies mentioned above envisage one – or at best a few – such forum(s), and thus limit participation to a very small number of individuals. Even if they were regularly held, the overwhelming majority of citizens would not gain a more active role in EU politics. Given that the opportunity for individuals to participate depends on the unlikely case of being randomly selected, one may also doubt that permanent citizens’ assemblies would contribute in a particularly significant way to the implementation of Article 10(3) TEU (the ‘participatory democracy article’), which provides that every citizen has the right to participate in the democratic life of the European Union.

The more important point, however, is that permanent citizens’ assemblies alone are not enough to achieve the change in the logic of representation that Landemore deems crucial for her version of participatory democracy. Landemore seeks to replace electoral mechanisms wherever possible to prevent the negative effects she attributes to them. By contrast, the envisaged permanent citizens’ assemblies would remain an add-on to the European Union’s (electoral-)representative apparatus. For the European Union to approach the model of open democracy, there would have to be far-reaching institutional reforms that fundamentally change the logic of the EU political system from an electoral to a primarily lottocratic one. Such reforms, however, are clearly not part of any of the proposals on the table. Permanent citizens’ assemblies as such, whatever their specific form, would not result in a profound transformation of the ordinary (representative) workings of the European Union. Against this background, it seems clear that participatory democracy is not a particularly fitting description for what they could achieve. Using this label would mean giving too much credit to what would ultimately be a small adjustment to the existing political system – and it carries the danger of raising expectations the European Union will not be able to live up to. We need to explain the democratic potential of permanent citizens’ assemblies in different – and more modest – terms. As I will argue now, this also allows us to hold on to, rather than to give up on, the idea of constituent power in the European Union.

V. Permanent citizens’ assemblies and the democratic minimum: Extracting a nucleus for constituent power

While the European Union is a democratic political system, it is a highly imperfect one – especially compared with the democratic orders of its member states (notorious counter-examples excluded – see Kelemen 2017). On the one hand, both the principles set out in Articles 9–12 TEU and the political institutions and practices for which they provide the basis – first of all, a directly elected supranational parliament – are undeniably of a democratic nature (Eriksen 2014: Ch. 3; von Bogdandy 2012). On the other hand, there is widespread agreement that EU democracy remains incomplete. For decades now, this view has been expressed with reference to the idea of a ‘democratic deficit’ at the EU level (e.g. Føllesdal and Hix 2006). The notion of a deficit plausibly suggests that the EU political system still lacks elements that are considered crucial for democracy at the national level – for example, a form of party competition that is politically consequential in the sense that it allows citizens to vote out the executive. Yet the image of a deficit is also misleading, as it suggests we know which ideal image the European Union is deficient towards. This is not the case. Even if most authors agree – given the European Union’s non-state character – that the point of EU democracy cannot be to replicate nation-state democracy (e.g. Nicolaïdis 2013), what exactly would be required for the European Union to become fully democratic remains hotly contested. Practically speaking, the European Union’s current state is that of a ‘laboratory for uncertain experiments’ with democracy beyond the state (Habermas 2019: 800, my translation). What EU democracy should ultimately look like is an open (political) question.

As a democratic project, the European Union can be interpreted as a response to the long-recognized fact that in a globalized world, many political problems exceed the regulatory capacities of nation-states. Growing interdependencies between states call for, and have de facto triggered, an extension of public authority – and, with it, ideally democracy – to the supra-state level. Two kinds of congruence problems drive this process. First, the scope of public authority must correspond to the political problems that need to be addressed; second, the boundaries of democratic inclusion must be (re)drawn to encompass those affected by political decisions (Habermas 2001; Zürn 2000). However, the transnationalization of public authority has been progressing faster than the transnationalization of democracy. While the European Union’s democratization may be somewhat more advanced than that of many forms of global governance, it is in the European Union that the most significant transfers of competences have taken place. Democracy still needs to catch up.⁸ In this context, it is useful to revisit Bohman’s idea of a democratic minimum. When he developed this notion, Bohman was concerned with how a political world characterized by cross-border problems and a corresponding proliferation of international institutions could be democratized – taking the European Union as the main empirical case. Bohman (2007: 47) argued that the path to democratization leads via a democratic minimum, which ‘attributes to each citizen the capacity to initiate deliberation and thus to take up the common activity of deliberating about common concerns, including the agenda of political institutions and the rules which guide political activity within them’.

Crucially, the democratic minimum does not include the power to bring about collective decisions. It falls short of (full) democracy. As Bohman (2007: 35) put it, the concept describes ‘a minimum of effective freedom, as opposed to the achievement of popular sovereignty in the sense of the final say over each and every decision’. The democratic minimum allows processes of democratization to get off the ground. The key to this is communicative freedom – defined by Bohman (2007: 8) as ‘the freedom to address others and be addressed as members of publics’. Members of different national political communities can recognize each other as participants in various transnational publics. In this way, they can establish contexts for deliberation about structures of public authority that affect them collectively. However, the power of informal publics is limited. To become ‘the potential source of communicative freedom and novelty’, transnational publics need to begin ‘to interact with and shape institutions’ (Bohman 2007: 14–15). According to Bohman, one way of supporting such processes is to set up mini-publics that, he argued with regard to the European Union, can serve to extend citizens’ exercise of communicative freedom in transnational public spheres to more formal settings; these, even if they have been established by regular political institutions, cannot be fully controlled by them:

By interacting with deliberative institutions at various levels, members of minipublics also interact with each other, thereby beginning a process of deepening democracy over which the delegating institution has no direct control. As empowered members of various polities and of the EU itself, such participants can make claims to other publics and to other institutions as they exercise their political rights as members of the European polity. (Bohman 2007: 88–89, emphasis added)

In short, the goal of communicative freedom in the context of less than fully democratic orders beyond the state, such as the European Union, is ‘to challenge and contest the new dispersed forms of delegated authority’ and to push for their democratization (Bohman 2007: 15).

The idea of a democratic minimum allows us to formulate a more plausible interpretation of the democratic potential of permanent citizens’ assemblies. In principle, EU citizens have the necessary political rights (in part due to national citizenship) to address each other as members of publics and to engage in the exercise of communicative freedom. However, it remains difficult for ordinary citizens, to say the least, to put policy issues on the EU agenda, let alone to raise fundamental questions of treaty change and institutional reforms. In Bohman’s (2007: 16) words, the European Union suffers from ‘a deliberative deficit, a deficit in the reflexive capacity of citizens to initiate democratic reform’. The point I make here is that the main purpose of permanent EU citizens’ assemblies could be to institutionalize a version of the democratic minimum in the European Union. If such an institution had a regular role in the decision-making processes of the European Union – that is, if its outputs could not be legitimately ignored but had to be systematically taken into account – this would establish a relevant channel for initiating deliberation about common concerns. A permanent citizens’ assembly could serve as a focal point and platform for initiatives of EU democratization that emerge from transnational publics and that, under present circumstances, have difficulties passing through the many institutional filters of the European Union’s ordinary political processes, many of which are characterized by intergovernmental domination.

Crucially, if we think about permanent citizens’ assemblies in terms of the democratic minimum, their primary value lies less in organizing participation in policy debates and the ordinary decision-making of the European Union than in creating a visible stage for questions of polity design. They can open up a space for the public deliberation of institutional reforms that the constituted powers would be reluctant to ever take up because they challenge the status quo and the existing allocation of competences. Even without decision-making powers, citizens’ assemblies such as those proposed by Alemanno and Citizens Take Over Europe have the potential to make a significant contribution in this regard. Namely, they would provide a way of tabling issues that is currently unavailable in the European Union. As a representative body, a permanent citizens’ assembly would command a pro tanto legitimacy that would give it a powerful voice difficult to dismiss by the European Union’s regular powers. Especially if such an institution were open to bottom-up initiatives, via an online platform or otherwise, it would establish a way for ordinary citizens to promote new directions in European integration. The same applies to ad hoc assemblies if they can be initiated from below, through the collection of signatures. In short, citizens’ assemblies would enable a more effective articulation of political demands that can otherwise be ignored all too easily by the EU institutions. Through such institutions, transnational publics could push more forcefully for constitutional change aimed at democratizing the European Union. This – not the advent of participatory democracy – is the potential (and more limited) democratic contribution of permanent citizens’ assemblies.

Finally, if we think about permanent EU citizens’ assemblies in terms of the democratic minimum, a link to constituent power emerges. One of the difficult questions for any democratic theory of constitutional politics is what can lead from the mere articulation to the (legitimate) exercise of constituent power. In principle, anyone who is subject to the public authority of constituted powers can enter the public sphere and articulate claims to constituent power. However, to trigger legitimate decisions on legal norms, a process of activation needs to follow in which those who articulate constituent power have to convince their political community to embark on a project of constitutional politics. If this succeeds, the actual exercise of constituent power then needs to happen according to legitimate procedures (Niesen 2019: 40–2; Patberg 2020: 182). In the European Union, this sequence is particularly difficult to initiate, not only due to the lack of procedures for the exercise of constituent power, but also because articulations of constituent power have a hard time finding adequate resonance in the first place due to the lower integration of the public sphere (as compared to national democracies) and the lower degree of organization of collective actors (e.g. a lack of truly European parties). Permanent citizens’ assemblies could provide a platform on which plausible bottom-up constitutional initiatives could progress from the articulation to the activation of constituent power. They could create institutional focal points around which public debate could develop and from which, eventually, demands for the exercise of constituent power – for example, for a permanent constitutional assembly – could be formulated. In this way, permanent citizens’ assemblies could become a nucleus for constituent power in the European Union.⁹

VI. Citizens’ assemblies after the CoFoE: Lessons concerning constituent power

There is no automatism, however, that turns citizens’ assemblies into base camps of constituent power – and any such attempt is likely to meet with resistance from constituted powers. Consider the way in which the CoFoE’s Conference Plenary and Executive Board have dealt with the following recommendation from the CoFoE’s Citizens’ Panel 2: ‘We recommend that the European Union reopens the discussion about the constitution of Europe with a view to creating a constitution informed by the citizens of the European Union. Citizens should be able to vote in the creation of such a constitution’ (CoFoE 2022: 138, emphasis added). Much could be said about this recommendation. In the first instance, one might be tempted to respond that it would be politically unwise to revive the notion of an EU constitution that is called ‘constitution’. But the reality, of course, is that the European Union has a constitution. The European Union’s primary law empowers and limits the exercise of public authority – that is, it works as a functional equivalent to national constitutions (Grimm 2016: 36). The question rightfully raised in the quoted recommendation is how citizens could take democratic control of the EU constitutional order and determine the future of European integration themselves. In short, we already witnessed an attempt to use EU citizens’ assemblies as a starting point for constituent power.

It is telling that this particular Citizens’ Panel recommendation – in contrast to many others – made its way into the final proposals of the Conference Plenary only in a very watered-down form, with any reference to a right to vote on a potential constitution erased from the text (CoFoE 2022: 83–4). There is no mention of the recommendation in the responses of Commission, EP, and Council to the CoFoE’s final report. While the EP has called upon the member states to set in motion a process of treaty revision – that is, to trigger a Convention in accordance with Article 48 TEU – the list of issues to be addressed by this Convention does not include any changes to how decisions on treaty revisions shall be made, let alone a right of citizens to vote on constitutional changes or to vote for extraordinary representatives that make such decisions (European Parliament 2022: 2–3). In other words, the EP’s call for treaty revision can hardly be interpreted as implementing the recommendation that EU citizens should have electoral power in a process of constitutional renewal. Moreover, as I am writing this, it seems rather unlikely that a Convention will be held at all. The reservations on the part of many member states are too great. So far, the Council has merely ‘taken note of’ the EP’s resolution in a preliminary technical assessment of the CoFoE’s final report (Council of the European Union 2022: 11). If the watering down and subsequent ignoring of this particular Citizens’ Panel recommendation is any indication how a permanent citizens’ assembly would be treated by the European Union’s constituted powers, it seems likely that they would try to keep citizens at arm’s length with regard to any attempt to reclaim constituent power.

Generally, we should expect EU institutions to strive for a model of citizens’ assemblies that does not seriously challenge their position. In the words of Gráinne de Búrca (2020: 145, emphasis added), written before the CoFoE:

There is every reason to doubt the willingness of EU institutions and of Member State governments to establish a citizens’ assembly intended to have real influence. Such ‘civil society’ consultations as there have been at the EU level have almost always been carefully managed, choreographed, and controlled by the Commission, and directed towards goals and policies already specified by the EU.

In line with this, the Commission’s communication on how it will follow up on the CoFoE’s final proposals – particularly on the call for permanent citizens’ assemblies – points in the direction of ad hoc mini-publics to be convened by the Commission at its convenience and on carefully pre-selected topics. In the future, ‘the Commission will enable Citizens’ Panels to deliberate and make recommendations ahead of certain key proposals’ (European Commission 2022a: 5). The Commission Work Programme 2023 specified that the ‘new generation of citizens’ panels will deliberate on …food waste, learning mobility and virtual worlds’ (European Commission 2022b: 4). Unsurprisingly, all of these processes concluded in the spring of 2023 without any notable public response. If citizens’ assemblies are employed in this way – that is, as forums for the deliberation of feel-good topics predetermined by the Commission – there is not much left of their potential to realize the democratic minimum in the European Union. They simply establish a façade of participation. What becomes clear here is that citizens’ assemblies are Janus-faced institutions. While in principle they can function as a nucleus for constituent power, they can just as well become a fig-leaf for its continued usurpation. In particular, they can provide the European Union’s constituted powers with a new argument – a form of participation to point to – to be used strategically to reject calls for more extensive citizen empowerment.

Much therefore depends on the exact design of EU citizens’ assemblies. While the details of this are beyond the scope of this article, the discussion of the democratic minimum allows me to make a few crucial points. Generally, EU citizens’ assemblies should be organized as spaces of politicization. Cristina Lafont (2020: 143) has distinguished three ‘political uses’ of mini-publics: contestatory, vigilant, and anticipatory. They all aim to improve the formation of considered public opinions in the context of ongoing policy debates. By contrast, a citizens’ assembly meant to institutionalize the democratic minimum has a different purpose – it must be constitutive. The goal is not to create a forum where matters can properly be discussed to the end, but to establish a platform from which deliberation reaching beyond the confines of the institution can be initiated effectively. This has two main implications. First, to realize the democratic minimum, the freedom to set the agenda is more important than rules for effective follow-up. An EU citizens’ assembly would need full autonomy vis-à-vis the EU institutions in deciding which topics to tackle. Nothing must be off limits. Only if the constituted powers are unable to prevent the deliberation of (from their point of view) unwelcome political ideas do citizens really enjoy the ‘capacity to begin’, as Bohman (2007: 52) called it, with a nod to Hannah Arendt.¹⁰ Second, an EU citizens’ assembly should be conceived of not as a mini-public, which merely simulates society-wide deliberation, but as an extension, or instrument, of the general public. Citizens’ assemblies can only fulfil their constitutive purpose if they are strongly embedded in the public sphere – that is, if there is a functioning feedback loop between members and non-members. This requires a wide distribution of agenda-setting power among ordinary citizens. Those who are not randomly selected into the assembly must be able to put items on the agenda (or to initiate ad hoc assemblies) and to provide input to the deliberation among members via mechanisms such as petitions or crowdsourcing. The democratic minimum must be established for all EU citizens.

VII. Conclusion

In this article, I have examined to what extent the Conference on the Future of Europe has advanced EU democracy. For a long time, the European Union has been shaped by constituted powers, contradicting Sieyès’s famous dictum that political institutions should not decide on their own terms of operation. In the months leading up to the Conference, one could get the idea that the process is an attempt to set things straight by finally empowering citizens to take the lead in determining the European Union’s future. Yet the CoFoE, especially the institutional setup of the Conference Plenary, fell short of anything that could be described as an exercise of constituent power. When this became clear, the democratic hopes invested in the Conference shifted their focus. Now the idea gained traction that the Citizens’ Panels held as part of the CoFoE could develop prefigurative power and lay the foundation for a form of EU participatory democracy that centres around permanent citizens’ assemblies. Against this, I have argued that the proposed institutions would not amount to participatory democracy in any theoretically demanding sense. Neither would they achieve the kind of societal transformation envisaged by Pateman (1970, 2012) and Barber (1984), nor would they amount to a clear departure from electoral democracy as sought by Landemore. The (potential) democratic contribution of permanent citizens’ assemblies needs to be explained differently. If we do so, we can also see that we should not yet give up on the idea that the CoFoE process might ultimately facilitate a reclaiming of constituent power in the European Union.

Specifically, I have argued that the institutionalization of EU citizens’ assemblies could help to realize the democratic minimum – that is, the capacity of citizens to initiate deliberation about common affairs, including the design of the EU polity. Here lies their (potential) democratic contribution. As a visible stage for bottom-up constitutional initiatives, EU citizens’ assemblies could enable transnational publics to move from the articulation to the activation of constituent power – and eventually even to push for mechanisms for its exercise. The problem – which is well-known from the domestic level – is that citizens’ assemblies also harbour considerable potential for democratic window-dressing and the rejection of more far-reaching demands for participation. The EU institutions’ response to the CoFoE and to calls for democratic constitution-making made from within one of the Citizens’ Panels has made this abundantly clear. Thus, whether EU citizens’ assemblies will really offer a nucleus for constituent power depends on their institutional design. If their purpose is meant to be constitutive, they must be designed as spaces of politicization. This requires full autonomy vis-à-vis the EU institutions when it comes to setting the agenda. At the same time, EU citizens’ assemblies need to be understood not as mini-publics but as extensions of the general public, which means that agenda-setting power should be distributed as widely as possible among ordinary citizens. While a citizens’ assembly with these features would not turn the European Union into a participatory democracy in the sense of models such as Landemore’s, it would provide a (deliberative) platform for citizens to challenge the European Union’s constituted powers.