Representing European citizens: Why a Citizens’ Assembly should complement the European Parliament

Why a European Citizens’ Assembly Should Replace Sortition with Liquid Democracy

Chiara Valsangiacomo (University College Dublin) and Christina Isabel Zuber (European University Institute and University of Konstanz)


Several previous contributions to this debate identify a dilemma in Kalypso Nicolaïdis’ thought-provoking proposal for a European Citizens’ Assembly (ECA) (Cristina Lafont & Nadia Urbinati; Richard Bellamy; Svenja Ahlhaus & Eva Schmidt). Nicolaïdis aims to promote a “pan-European participatory ecosystem” characterised by a transnational “demoicratic ethos.” She argues that the ECA could foster these goals, enhancing input, throughput and output legitimacy of collective decision-making within the EU. As critics point out, for the ECA to have such an impact on the quality of transnational democracy, it would have to be equipped with some degree of decision-making authority. However, if it did have such authority, it should not rely on selection by lottery. These critical contributions solve the dilemma by downgrading the ECA to a merely consultative body, arguing that we should strengthen traditional forms of electoral democracy instead. We support both Nicolaïdis’ ambition to radically reimagine EU institutions beyond elections and the arguments against lottocracy. Our contribution, therefore, suggests a different way forward, one that Jelena Dzankic briefly hints at: replacing selection by lottery with liquid democracy.

The Problem with Sortition

Faced with citizens who neither care about nor feel represented by the institutions of the EU, Nicolaïdis suggests that we shift gears and consider “radical institutional innovation”. Her specific proposal is to complement the European Parliament (EP) with a deliberative ECA, whose members would be drawn by lot from all 27 states and rotate on a regular basis. Dzankic emphasises the underspecified nature of this proposal. What does seem clear, however, is that Nicolaïdis envisions the ECA as an institution that is permanently embedded in the EU’s institutional regime. With an independent position and budget, it would contribute to legislative agenda-setting, deliberate on policy issues and prepare European Citizens’ Initiatives (ECIs), work closely with the EP, monitor policy implementation, scrutinise other EU institutions, and cooperate with civil society actors. According to Nicolaïdis, such an assembly would enjoy input, throughput, and output legitimacy because it would give citizens an equal chance to be selected, help combat problems of corruption and elite capture that bedevil electoral institutions and foster epistemically valuable forms of collective intelligence.

Nicolaïdis’ discusses several ways in which the ECA could connect to the EU’s wider public sphere and help citizens come to feel that they own collective decisions made at the European level: mixing ECA members with professional, elected members of the EP, coupling the ECA with direct democracy, embedding it in civil society, crowdsourcing, fostering media partnerships, and giving the ECA an itinerant, traveling nature.

Nicolaïdis is cautious to not grant the ECA binding decision-making power on its own, but foresees it playing a co-decision role in the EU’s power-sharing system. As Cristina Lafont and Nadia Urbinati point out, in case of conflict over the substance of decisions between the ECA and other European institutions, there would need to be a rule on who prevails. For the ECA to genuinely play the role of increasing input, throughput and output legitimacy that Nicolaïdis hopes for, it would need some kind of binding decision-making authority. However, if it did have such authority, it should not be selected by lottery. As Bellamy, among several other contributors, points out, sortition stands in the way of overcoming the present “demoi-cratic disconnect”. In particular, we share Lafont’s &  Urbinati’s diagnosis that a lottocratic ECA must, by design, rely on the “blind deference” of the vast majority of citizens to the decisions of a small group of randomly selected – and thus democratically unaccountable – individuals. Since the wider public has “no role to play in the functioning of these institutions,” low participation and disinterest are bound to remain the norm. As long as citizens are denied genuine agency (Bellamy), the lottocratic ECA is bound to remain just another “shortcut” (Lafont & Urbinati) that cannot justify a sense of “democratic ownership” (Ahlhaus & Schmidt).

The solutions offered by Nicolaïdis’ critics by and large aim at preserving the benefits of lotteries for selecting the ECA’s members. They therefore opt to downgrade the role of the ECA from collective decision-making to mediation or consultation. Lafont & Urbinati propose an intermediary ECA that is not “directly coupled with formal political institutions”. This idea is welcomed by Sandra Seubert as well as Ahlhaus & Schmidt. Others prefer to stick to temporary and issue-specific assemblies, similar to earlier citizens’ panels (Daniel Freund, Graham Smith & David Owen). What all of these proposals have in common is that they ultimately abandon Nicolaïdis’ ambition to give ordinary citizens the power to participate more directly in the EU’s legislative process. By continuing to limit the exercise of democratic sovereignty to elected representatives, they can offer no solution to the challenges of elite capture and wide-spread citizen disengagement. Because we support both Nicolaïdis’ ambition to radically reimagine EU demoicracy beyond elections and the arguments against lottocracy, we want to suggest a non-lottocratic, non-electoral alternative: a so-called liquid ECA. This innovation would perform the same institutional functions as Nicolaïdis’ lottocratic ECA, but the selection of its participants would be based on a completely different logic, namely liquid democracy.

A Liquid European Citizens’ Assembly

Liquid democracy is a technology-enabled mechanism for distributing voting power that has recently gained attention in several disciplines, including computer science and political theory (see, e.g., Halpern et al., Ford, Landemore, Valsangiacomo). It came to prominence in the 2010s thanks to the experience of the German Pirate Party, and it can be technically implemented with software such as LiquidFeedback or Adhocracy+ (see, e.g., Behrens et al., Deseriis, Paulin, Valsangiacomo). The liquid ECA we envision would have to incorporate all the defining features of liquid democracy: “direct democracy, flexible delegation, meta-delegation, and instant recall” (Blum and Zuber). For each decision, members of the political community can either vote directly or delegate their vote to another citizen. They are free to recall these delegations at any time. The delegates are called “proxies”, and they cast as many votes as they received, plus their own. Proxies can also meta-delegate votes to other proxies, creating a transitive chain of delegations. Because liquid democracy does not limit the total number of active participants, the group of decision-makers can potentially encompass all members of the political community. Moreover, even proxies with few delegations are allowed to act as representatives, creating a perfectly proportional system. For these reasons, liquid democracy is a particularly inclusive system with low barriers to political participation. It is also a remarkably flexible decision-making system in which patterns of participation and delegation can change quickly, depending on the issue at hand. Thanks to policy-area-specific delegation, our liquid ECA would resemble Alexander Guerrero‘s single-issue legislatures: it would consist of a predetermined number of independent and autonomous sub-assemblies (for example, mirroring the standing committees of the EP), so that an EU citizen or resident could be actively involved in legislation on fishery, while being represented by different proxies in the area of trade policy, or on matters concerning the single market.

According to Blum and Zuber, liquid democracy is superior to electoral democracy in realising popular sovereignty or input legitimacy. If the lottocratic ECA were transformed into a liquid ECA, it could be given genuine legislative authority without undermining the principle of democratic sovereignty: each member of the political community would either participate in decision-making directly or authorise representatives, thereby avoiding blind deference. A liquid ECA would thus allow all citizens to actually share in political power, overcoming the “vertical inequality” between a group of active decision-makers and a passive group of ordinary members that otherwise persists in both electoral and lottocratic systems (Abizadeh). This would foster a “justified sense” of democratic ownership (Ahlhaus & Schmidt). At the same time, a liquid ECA would also offer strong procedural and epistemic advantages (respectively, throughput and output legitimacy) over the EP or Nicolaïdis’ model for an ECA, thanks to the possibility of delegating to policy experts, rather than electing generalist representatives or deferring to randomly selected individuals, neither of whom may have the relevant expertise. First, the task of monitoring representatives’ behavior is reduced to specific policy fields, lowering the burden of ensuring meaningful accountability and limiting the possibilities of capture (a problem discussed by Seubert). Second, area-specific delegation and low barriers to participation help identify and activate individuals with the relevant policy-specific expertise to participate in legislative decision-making, mobilising the wisdom of the crowd (Blum and Zuber, for experimental empirical support see Revel et al.).

Choosing a liquid over a lottocratic ECA almost inevitably requires the extensive use of digital technologies to work, moving the assembly closer to a situation where politics is done largely online. This new virtual dimension could provide additional advantages. First, policy-specific deliberations in the liquid ECA could draw on the power of AI-assisted translation, allowing EU citizens and residents to participate regardless of their language skills and physical location (thereby addressing concerns raised by Dzankic, Bauböck, and Smith & Owen). Second, the policy-area-specific nature of deliberation in the liquid sub-assemblies could foster the development of non-territorial transnational communities around policy issues, complementing the territorial and ideological logics of representation that currently dominate in the EP.

A Thought Experiment

We are aware that while deliberative mini-publics have become more popular in recent years and have already been tried from the local to the European level, experiments with liquid democracy have so far been limited to decision-making processes within political parties (see, e.g., Kling et al., Adler). However, deliberative mini-publics were also once mainly discussed by democratic theorists rather than applied in practice.

Let us, therefore, conclude with a thought experiment to make our proposal more plausible and tangible (for a similar approach, see Landemore). Imagine we were to equip each and every EU citizen or resident with an account on a pan-European version of a liquid feedback software. Once logged in, citizens could see a visual representation of the structure of the liquid ECA, with its independent, thematic sub-assemblies, and decide in which policy areas they wish to participate directly and in which to delegate. Next, whenever action is required by one of these sub-assemblies, all the proxies and citizens who have registered as active participants for that specific policy area would be notified and invited to participate in online deliberations and, eventually, vote on the issue at stake based on liquid democratic principles. Anyone who wanted to have a direct say on the matter would have had the chance to bring their perspective and interests into the discussion, anyone who preferred to be represented by someone they trust on that policy area would not participate and be represented by proxy instead. Would this not open up many more “ways of channelling the life wisdom, knowledge spheres and expertise of a broader range of individuals than those self-selected in the political and bureaucratic spheres” (Nicolaïdis) than deliberations among a small group of randomly selected individuals ever could?