STUDY DEMOCRACY AND HUMAN RIGHTS CITIZENS’ ASSEMBLIES New Ways to Democratize Democracy Wolfgang Merkel, Filip Mila č i ć and Andreas Schäfer September 2021 Citizens’ assemblies can ­provide a constructive response to the current erosive symptoms of re­ presentative democracies by injecting the perspec­­tives of ordinary citizens into the political process. Citizens’ assemblies must fol­ low a precise design. The cru­ cial criteria are: representative inclusion, high-quality delib­ erative procedures, and sys­ temic efficacy. They must have a binding, transparent goal, yet not specify in ad­ vance what the outcome of their deliberations should be. Citizens’ assemblies are no substitute for the parties and institutions of representative democracy. But they can bring the latter closer to the citizenry and foster public will-formation.

CITIZENS’ ASSEMBLIES New Ways to Democratize Democracy Citizens’ assemblies can offer a construc­ tive response to those challenges current­ ly plaguing representative democracies such as loss of trust, selective non-partic­ ipation, and social polarization. But they will not be able to realize this potential unless they meet certain criteria: repre­ sentative inclusion, high-quality delibera­ tive procedures, and systemic efficacy. The procedure for drawing lots must be arranged in such a way that people from all affected social strata take part. Also, the internal communication process must be organized and facilitated so that all of the participants can get involved in the process as equals, regardless of their so­ cial background or habitus. The process must be open to a variety of outcomes; and citizens’ assemblies must be embed­ ded in the democratic political system as autonomous and influential factors. In short, the citizens’ assemblies must follow a precise design that also should feature a binding, transparent set of targets. Thus, citizens’ assemblies have a good chance of living up to the normative expectations placed on them and of bringing the per­ spectives of“ordinary citizens” effectively into the political process. In those ways citizens’ assemblies can help narrow the chasm that separates cit­ izens and political decision-makers. They introduce a new logic into representative systems currently dominated by electoral competition. It may enable us to avoid a variety of dysfunctional phenomena: the undue focus on elites, class-specific ab­ stention from politics, the inflated signifi­ cance of political marketing at the ex­ pense of substance, and the infatuation of democratic politics with present rather than future issues. When citizens and the broader public realize that citizens’ as­ semblies keep the promises they make, trust in democratic institutions will re­ bound, public will-formation should be encouraged, and participation in the po­ litical process will be both widened and deepened. Further information on the topic can be found here: democracy.fes.de

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES DEMOCRACY AND HUMAN RIGHTS CITIZENS’ ASSEMBLIES New Ways to Democratize Democracy B

 Content 1 INTRODUCTION: BACKGROUND AND A DIAGNOSIS OF THE PROBLEM  2 1.1 The 4 1.2 The 5 1.3 The 5 2 THE DISCUSSION IN DEMOCRATIC THEORY 2.1 2.2 Normative Principles and 2.3 Systemic 9 2.3 Interim 10 3 POLITICAL EXPERIENCES AND EMPIRICAL FINDINGS  11 3.1 11 3.2 A Comparison of Paradigm 3.3 The Balance 3.4 Unresolved Questions and  DESIGN OF CITIZENS’ ASSEMBLIES   18  Basic Principles: Flexibility in Design, Binding and Transparent 18 4.2 Organization and 19 4.3 Selection and Recruitment of Participants 20 4.4 Process 21  and Leveling of Structural 23 4.6 Media 24  Control, and 24  Efficacy and Compatibility with Representative 5 CONCLUSION AND OUTLOOK FOR THE FUTURE  28 How to Make a Citizens’ Assembly Succeed  31 Interview 35

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES 1 INTRODUCTION: BACKGROUND AND A DIAGNOSIS OF THE PROBLEM How is our democracy really doing? We hear that question asked with disquieting frequency. The very fact that we hear it so often may be taken as a sign that it is not doing very well at all. If one consults recent academic studies of democracy, there is an unambiguous, dominant theme: Yes, democracy(assumed to be a singular noun) is in deep trouble. Titles like The Crises of Democracy(Ralf Dahren­ dorf), Existential Crisis of Democracy(Jens Hacke), The People vs. Democracy: Why Our Freedom Is in Danger and How to Save It(Yasha Mounk), The End of Democracy (Yvonne Hofstätter, confidently), and — more circumspect­ ly —  Democracy and Crisis(Wolfgang Merkel and Sascha Kneip, 2018) reinforce that impression. Every major Europe­ an language can boast a series of books with similar“crisis titles.” In English they are of course quite common, as ex­ emplified by: Life and Death of Democracy(John Keane), How Democracies Die(Steven Levitsky/Daniel Ziblatt), Democracy in Decline(Philipp Kottler), not to mention many entitled Democracy in Crisis or, more succinctly, Democrisis (David Roche). We will begin this study on a more cautious note. We will not simply assume that there is a crisis of democracy, a de­ cision for which both theoretical and empirical grounds can be cited. – First, it is misleading to talk about a“crisis of democ­ racy” in the singular, as though democracy were monolithic. Democracy in the USA is not the same thing as it is in Canada, while the Danish variant is distinct from its Austrian counterpart. Germany is not Poland, nor is France Hungary. In contrast to political theory, empirical research on democracy always must keep such distinctions in mind — indeed sharpen them in order to make comparisons and thus track down the causes behind positive and negative trends. This must be done not only on a global scale but even within the EU. Hungary’s democracy, especially, has been defective for quite a while, and that of Poland increasingly so; consequently, both really are in the midst of a crisis. 1 Yet Sweden, Finland, and Germany are not. But this observation does not mean that the older, more consolidated democracies do not face unre­ solved challenges. Those include, among others, the low and/or vanishing trust of the citizenry in the or­ ganizations(parties) and institutions(parliaments, gov­ ernments) that are supposed to represent them; the gap between acceptance of democratic values and the perceived performance of really existing democracies; the rise of right-wing populist parties; the revival of ethnic nationalism; growing social and political polari­ zation 2 ; as well as socio-economic inequality, still a po­ litically potent issue. All affect the participatory, egali­ tarian, and deliberative principles of democracy and vi­ tiate its quality. – Second, we lack robust theories of democracy that might provide us with a precise selection of adequate indicators and quantitative threshold values that would enable us to measure exactly at what point a democ­ racy enters or emerges from the crisis phase. For exam­ ple, if the far-right Alternative for Germany(AfD) in Germany were to win 20 percent of the vote, would that be enough to allow us to diagnose that there was a crisis of democracy(in the singular)? Or would it sig­ nify a crisis if the AfD entered government as a coali­ tion partner? In Austria, Denmark, Switzerland, and France(in presidential elections) right-wing populists already have racked up even higher vote totals. The populist right already has participated in coalitions in more than ten European governments, including those of the Scandinavian countries which always are ac­ corded the highest scores for the quality of their de­ mocracies( Varieties of Democracies; Freedom House; Democracy Barometer). Would a decline of 20 percent in voter turnout in Germany justify us in talking about a crisis of democracy there? If the answer is yes, then what argument should we make when we recognize that many other democracies have even lower voter turnout rates, including Switzerland, the USA in its midterm Congressional elections, or Poland in its par­ liamentary elections? In most of those cases, turnout usually falls below 50 percent. Here, too, the list of examples could be extended. 1 For more on the“difficult” concept of the crisis of democracy, cf. Merkel 2016 et al. 2 On the notion of polarization cf. Merkel 2021. 2

Introduction: background and a diagnosis of the problem Figure 1 The development of democracy in EU countries and established democracies(1950–2020)  3 85% 80% 75% 70% 65% 60% 55% 50% 45% 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 Source: V Dem—Varieties of Democracy: Average of the 26 EU member states(excluding Malta), the United Kingdom, Australia, New Zealand, Canada, the USA, and Japan; https://www.v-dem.net/en/data/data/ – Third, no single indicator suffices to enable us to diag­ nose the presence of a crisis. For decades now, we have experienced the asynchronous development of different organizations, institutions, and procedures among democracies. While some are ascending, oth­ ers are in decline. Thus, LGBTQ people, same-sex cou­ ples, ethnic minorities, and women now justifiably en­ joy considerably more rights and less discrimination than they did 10, 20, 30, or 50 years ago. But at the same time, in the Europe of five decades ago it was easier to steer the“national economy” toward demo­ cratic policy goals than it is today, in the age of globali­ zation. For these and other reasons(Merkel 2015: p. 7 ff.) we will speak of unresolved challenges among established democ­ racies, although admittedly those challenges represent sig­ nificant backsliding in the quality of developed democracies over the last ten years(Merkel/Lührmann 2021). Precisely because of these continuing trends, we will speak of the erosion of most(established) democracies. Erosion here does not refer to a powerful, short-term, concentrated dip, but rather to a decades-long trend that has slowly under­ mined and weakened democracy. The graph below — which depicts the quality of democracy in 27 EU countries(includ­ ing Great Britain but excluding Malta) as well as in some other established democracies(Australia, New Zealand, Ja­ pan, Canada, and the USA) — clearly registers this drop in quality beginning in 2010, even though the political re­ gimes featured here on average had chalked up steady, steep qualitative gains since 1950. In addition to the heavily aggregated trend-line of democ­ racy among 32 countries, we have chosen four that again reveal the“erosion of democracy” in exemplary fashion. The experts who prepared Varieties of Democracy(V-Dem) picked out Denmark from among 200 countries as having the best democracy. But later on, even Denmark experi­ enced a slight deterioration in quality. The regression was significantly greater in Germany and Austria. The United States, which needed 25 more years after 1950 to match Western Europe’s standard of democracy, fell off a cliff af­ ter 2016, when its 45th president held office. Even without relying on a precise theory of democracy to determine the quantitative threshold values, one can say that, under Don­ ald Trump, the USA went through a tumultuous crisis in a compressed span of time. 3 Of course, the countries of Eastern Europe were authoritarian re­ gimes until 1990 and democracies only after that. But even author­ itarian regimes can have some democratic features, and those, too, were incorporated into the findings on Eastern Europe for the pe­ riod from 1950–1990. 3

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES Figure 2 Development of Democracy in Denmark, Germany, Austria, and the USA(1950–2020) 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 Denmark Germany___ Austria__ USA Source: V Dem—Varieties of Democracy 2020; https://www.v-dem.net/en/data/data/ This is not the place to identify and detail all the phenome­ na associated with the erosion of democracy or to pin down its causes(Lührmann/Merkel 2021). Instead, we will take a brief glance at the problems of democracy linked to political participation. That strategy makes sense because we will be looking at the prospects for democracy latent in citizens’ as­ semblies and those are mostly to be found in civic participa­ tion. The issue we will examine concerns whether citizens’ assemblies contribute to counterbalancing the deficits of democracy, and, if so, in what form they might do so best? In other words, what potential for(re-)democratization do citizens’ assemblies harbor, and what roadblocks stand in the way of their realizing that potential? 1.1 THE MICRO-LEVEL the favor of voters through broad-gauged marketing strategies that, in turn, eventually force citizens into the position of consumers. Meanwhile, as parties embrace marketing and forfeit the loyalty of their voters, they find it harder to aggregate interests and viewpoints, which is one of their key functions in a democracy. – Social Selectivity. Citizens who take advantage of op­ portunities to participate in politics are predominantly those with formal education and/or those who are in­ terested in politics. The bottom third on the education scale mostly have stopped participating; moreover, they perceive themselves as having low levels of politi­ cal efficacy(Schäfer 2010). As far as participation is concerned, Germany—like most other member states of the EU—can be labeled a high-functioning“twothirds democracy.” – Declining party affiliation and volatile voting behavior. In recent years the volatility of voting behavior has in­ creased considerably. Although that reflects a certain dissatisfaction with parties generally, it also can be in­ terpreted as an indication of the“rational maturity” of enlightened citizens, who constantly weigh the perfor­ mance of those who govern them against their own interests and values. Decreasing party affiliation offers heightened incentives for actors in party politics to court – Polarization. Being“fed up with politics,” the general decline in voter turnout, and flagging political partici­ pation are by no means the“crisis signature” of our political epoch any longer(Meiering/Schäfer 2020). A new kind of social and political polarization has emerged to replace those symptoms. Confrontational pluralism pitting political opponents against each oth­ er, which once enlivened political conflicts, increasing­ ly is evolving into political polarization in which hostile 4

Introduction: background and a diagnosis of the problem camps face off. It is almost as though that controver­ sial scholar of constitutional law, Carl Schmitt, had written the political script. This kind of toxic polariza­ tion blocks compromise and imperils social cohesion. The last three major crises in Europe and beyond its boundaries—the migration crisis, the COVID-19 pan­ demic, and the climate crisis—have done much to ag­ gravate such polarization. 1.2 THE MESO-LEVEL –“Bowling alone.” Ever fewer people are attracted to collective organizations like parties, labor unions, and associations or clubs. One scholar of civil society, Rob­ ert Putnam(1995), memorably dubbed this withdraw­ al from social organizations“bowling alone.” Political parties have been especially hard-hit by the flight from associations(van Biezen/Poguntke 2014). In fewer than two decades they often have lost more than half of their membership. Worse still, parties are among the political institutions all across Europe that enjoy an especially low level of trust, as data from Eurobarome­ ter annually confirm anew. – Decline of integrative parties.“Catch-all” parties have been swept up in an irrevocable decline. They have lost votes, members, trust, and attractiveness. Besides, their membership has become hopelessly geriatric. They have disappeared in many countries and will no longer play a central role in the individualized societies of the 21 st century. That is a problem because, in the age of polarization, political integration machines(which is what the catch-all parties were in the second half of the twentieth century) have gone missing precisely when their integrative services are most needed. Smaller par­ ties with narrower programs will not be able to play this integrative role. – The rise of right-wing populism. On the right fringe of the political spectrum, right-wing populist parties have risen to become relevant political actors in almost all European party systems. These“semi-loyal parties” (Juan Linz) have an instrumental, illiberal understand­ ing of democracy. Their nationalistic identity politics worsens social and political polarization. 1.3 THE MACRO-LEVEL – Loss of trust. Both governments and parliaments have forfeited trust. That is true especially of the most im­ portant institution of representative democracy, the parliament. In the wake of the ongoing internationali­ zation of political decision-making, parliaments have been forced to surrender many of their remits to na­ tional and supra- or transnational executives. The fi­ nancial crisis of 2008, the migration crisis of 2015, and most notably the coronavirus crisis(2020 ff.) all have further accelerated this loss of substantial political re­ sponsibilities. – Decreasing representativeness. The power of parlia­ ments to represent citizens, and thus their credibility, have decreased. On one level, that decline affects socalled descriptive representation, i.e., the extent to which the representatives’ social and gender character­ istics correspond to or mirror those of the people they represent. The dominance of academics in parliament also finds expression in so-called substantive representa­ tion, i.e., the representation of interests. Here, too, the values of middle-class people with academic credentials dominate. In addition, empirical research shows that the preferences of the(upper) middle classes also are over­ represented in laws passed and policies adopted. The preferences of the lower classes remain underrepresent­ ed(Lehmann/Regel/Schlote 2015: p. 157 ff.). – Reduced cohesion. The cohesion of the political com­ munity has diminished; moreover, that decline is re­ flected in the renewed division of the societies of de­ veloped democracies into classes: upper and lower, or cosmopolitan and nationalist-communitarian camps. This poses a crucial test of the citizens’ identification with“their” commonwealth. People increasingly feel as though they belong only to their own class, their own camp, or their own familiar communities of value. Traditional political institutions and actors so far have not been able to build bridges in a society increasingly marked by segmentation. That difficulty in bridge-build­ ing often is aggravated(rather than reduced, as many had sometimes hoped) by the possibilities of online communication through social media. These are the most significant deficits and unsolved prob­ lems in the societies and political systems of established de­ mocracies. Being“fed up with politics” is the wrong way to describe the condition of our democracies. It is not merely polarization that augurs differently for us. One may also ob­ serve that there is a growing political—even democratic— demand among the people for more direct and active polit­ ical participation in their democratic communities. Numer­ ous digital platforms for discussions, exchanges, signature collecting drives, initiatives for referenda, and petitions or mobilizations for protests and demonstrations illustrate this as well. Even though not all such initiatives are run by pro-democracy forces, they show that many citizens want to do more than just vote. This suggests what empirical studies also confirm: that evaporating trust in institutions does not signify disillusionment about democracy and its principles; rather it seems to be influenced by disappointment about the performance of the institutions and procedures of rep­ resentative democracy. Based on experiences in Switzer­ land, we also know that direct citizen involvement above and beyond voting helps strengthen their identification with the political community. So, the question needs to be asked: which forms of political participation are best suited to(re-) democratizing our democracies? The Swiss model of popular plebiscites at all three territorial levels of the state cannot simply be transferred to other soci­ 5

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES eties. Generally speaking, models always need to be contex­ tualized. Although we would not want to reject popular ref­ erenda in general(Merkel/Ritzi 2017), caution may be in or­ der when it comes to holding them in polarized societies. Such plebiscites are zero-sum games. Whichever group man­ ages to gain more than 50 percent of the needed votes for its proposed resolution is the winner. Moreover, plebiscites can have an exclusionary character. But in segmented-polar­ ized societies, democratic innovations usually should build in some mechanisms of inclusion as pro-democracy antidotes to such exclusionary tendencies. As many people as possible should identify with the reformed institution and, as far as possible, desire to and be able to take part in it. Deliberative institutions obey a fundamentally different log­ ic than plebiscites. They are not about quantitative-numeric results designed to generate“decisionistic” outcomes. Rath­ er, mutual and empathetic exchanges of arguments are their central feature(Schäfer/Merkel 2020). Ideally, what counts is only the“peculiarly unforced force of the better ar­ gument,” as Jürgen Habermas put it succinctly. A constitu­ tive element of citizens’ assemblies is the deliberative meth­ od: that is, an approach that, in the current condition of our society, suggests itself as a kind of practice that can build bridges across the trenches. But it does not get us very far merely to offer such an assessment of broad principles. Many other aspects must be examined: How inclusive are citizens’ assemblies? Who participates in them? How can they be combined with representative institutions like parlia­ ment? Which design of citizens’ assemblies strengthens their democratizing effects? What kinds of rules produce problematic consequences? Those issues will be addressed in the following pages. 6

The discussion in democratic theory 2 THE DISCUSSION IN DEMOCRATIC THEORY 2.1 DEFINITIONS Citizens’ assemblies represent a specialized form of so-called “mini-publics” first conceptualized by Robert Dahl, a scholar of democracy. To Dahl, they suggested a way to solve a spe­ cific problem: In the process of political decision-making, an ever-wider chasm has opened up between political elites and average citizens(the demos). Because the latter lack re­ sources such as knowledge and attention, they seem less and less able to judge complex matters that figure into deci­ sion-making(Dahl 1989, p. 338 ff.). In his view, a“mini-­ populus” might help to bridge this gap by generating from among the entire demos a broadly representative — albeit relatively manageable — group through random selection. Representing the entire people collectively, it would deliber­ ate, along with experts, on a political issue within a limited time-span, and — based on those deliberations — would adopt a public position. In other words, mini-publics are suf­ ficiently small to make possible an interactive discourse, but large enough that they could constitute a representative sample of the population(Goodin/Dryzek 2006, p. 220). What all forms of mini-publics have in common is the use of random selection to recruit participants(Smith 2009). In this respect they differ both from classical forms of participation as well as from other democratic innovations such as partic­ ipatory budgeting, in which all members are self-selected. Of course, a person’s decision about whether or not to par­ ticipate also plays some role in mini-publics, because as a general rule participation is voluntary, which means that a certain significance may accrue to self-selection(cf. section 3). Nevertheless, only those who have been previously cho­ sen by lot and invited may take part. Furthermore, the invi­ tation to participate is valid exclusively for the concrete, time-limited format. Moreover, as compared to other forms the procedure itself is rather strongly(pre-) structured and facilitated, the goal being to enable the most inclusive and intensive deliberative process possible among participants. Citizens’ assemblies generally differ from other forms of mi­ ni-publics in terms of their size, their objectives, and the time-span during which they take place. With between 100– 160 participants, citizens’ assemblies exceed in numbers citi­ zens’ juries for instance(typically 12–26) or“consensus con­ ferences”(roughly 10–18). In contrast to“deliberative polls,” they aim not to ascertain political opinions, but to issue con­ crete political recommendations. Also, with some 20–30 ses­ sion days stretching over several weeks or even months, they last for a relatively long time(Elstub 2014). The Citizens’ As­ sembly of British Columbia, which deliberated on a reform of the electoral law in that Canadian province in 2004, is con­ sidered to be the first(modern) instance of this format. Of course, the boundaries between citizens’ assemblies and other mini-publics are fluid. As far as structures and goals are concerned, they display a correspondingly wide variety of options for designing, varying, and combining different ele­ ments. So, it is all the more important to reflect on and har­ monize goals and forms in the process of concrete imple­ mentation(cf. chapter 4). 2.2 ORMATIVE PRINCIPLES AND EXPECTATIONS Theories of participatory and deliberative democracy pro­ vide the normative foundation for the design of and critical reflection upon citizens’ assemblies(Barber 1994; Chambers 2003; Habermas 1994; Dryzek 2000). Accordingly, delibera­ tive participation is the general guideline for the format of this democratic innovation. By deliberation we mean the in­ teractive exchange and collective weighing and balancing of arguments in view of a problem calling for a decision to be made. Thus, it is to be expected that the formation of opin­ ions and development of positions will come about through argument, information-assimilation, and learning processes, as the case may be. To some extent this process can be anal­ ogized to parliamentary deliberative procedures(Schäfer 2017). Nevertheless, what is envisaged here is the participa­ tion of citizens from the most diverse possible strata and mi­ lieus of the population. The theory of deliberative democra­ cy furnishes a model of legitimation for will-formation and decision-making. Its normative expectations in regard to the legitimacy of citizens’ assemblies can be further broken down into three aspects: recruitment, process design, and outcome. THE RECRUITMENT PROCESS The recruitment process for citizens’ assemblies makes use of both sources of legitimacy noted above. For one thing, random selection by lot can be regarded as the democratic 7

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES procedure par excellence(Buchstein 2009; Manin 1997), be­ cause it absolutely ensures equality among all citizens. This is the case since, in principle, the lottery scheme gives every citizen the same chance to be chosen and enjoy the oppor­ tunity to participate. Anyone whose name was not drawn by lot can at least be certain that the selection was fair and treated all eligible people the same regardless of status, and that on future occasions they will have the same chance to be chosen. In the ancient world’s precursors to our citizens’ assemblies, the drawing of lots played a crucial role, particu­ larly in the Athenian councils. Most political offices in an­ cient democracies were filled by that method. However, in addition to the equality of opportunity guaranteed by the drawing of lots, there is a second source of legitimacy inher­ ent in the practice of random selection. It aspires to ensure that participating groups of citizens are statistically repre­ sentative of the population. The composition of a randomly selected body should mirror the heterogeneity of the entire population’s demographic profile. So, ideally, the selected body will be a group of average people. Thus, anyone not chosen should still suppose that people“like you and me” will be represented on the citizens’ assembly. Against this background it is thus plausible for non-participants to as­ sume that, if they had been involved personally in the delib­ erations, they would have ended up taking into account in­ terests and points of view similar to the ones actually con­ sidered. Since the recruitment procedure for citizens’ assem­ blies makes use of both principles of legitimacy, it combines both representative and direct democracy. Occasionally, however, there can be good reasons to deviate from the principle of pure random selection. In practice “stratified” random selection often is adopted to make sure that all relevant social perspectives on the issue at hand will be represented in the group of participants(Farrell/Stone 2020: Ryan/Smith 2014). To cite one example, it may turn out that certain minorities or regional perspectives will be statisti­ cally overrepresented, but for legitimate reasons. Also, de­ pending on the specific cases or matters to be decided, it may seem appropriate to modify the population from which par­ ticipants are randomly selected. This might happen if some groups within the population would be especially deeply af­ fected by the decision reached, or if the distribution of inter­ est in participating was not equal across the board. Some ap­ proaches take a further step in that direction by suggesting that politically disadvantaged groups be given the opportuni­ ty — at first in relatively homogeneous mini-publics — to de­ velop their own perspectives in a collective exchange of views. In this phase of“enclave deliberation”(Karpowitz et al. 2009) representatives of such groups chosen by lot could develop both the substantive arguments and positions and the necessary self-confidence to take part on an equal foot­ ing in forums with a socially heterogeneous make-up. DELIBERATIVE PROCESS A twofold normative expectation is associated with deliber­ ative processes in the framework of citizens’ assemblies as well as more generally: inclusion and“epistemic fruitfulness” (Bächtiger et al. 2014). First, every participant should be able to take part effectively and on an equal footing with others. Second, the open, critical, and respectful consideration of the most diverse perspectives should encourage a high-qual­ ity discourse. Participants should be enabled to develop, jus­ tify, and reflect upon both their own and commonly shared positions in the course of discursive debates and the give and take of reasoning. Some empirically well-founded hopes are associated with the deliberative process: that the mutual un­ derstanding of opposing arguments might increase, poten­ tial prejudices might be overcome in favor of learning ef­ fects, while problematic group effects such as the pressure to conform could be circumvented(List et al. 2013). To fulfill those hopes, certain steps may be regarded as crucial such as making available a well-balanced supply of information relevant to the decision, ensuring that the make-up of the group is sufficiently diverse, and facilitating exchanges in small group discussions aimed at attaining the most inclusive possible participation of all the members(Setälä/Smith 2018). The first cases of citizens’ assemblies in British Colum­ bia, Ontario, and the Netherlands were constructed in a sim­ ilar manner(Fournier et al. 2011: pp. 21–50). Right after the selection process was complete, a learning phase followed, in which information on the different electoral systems on which the citizens’ assembly was supposed to develop a po­ sition was provided by neutral experts. After that, a consultation phase ensued in which members of the broader public could make their own perspectives known to the as­ sembly members through regional public hearings and de­ bates. The process concluded with an internal deliberation and decision-making phase, in which plenary meetings al­ ternated with small group sessions moderated by trained personnel. THE OUTCOME Against the backdrop of these notions of how deliberative mini-publics are supposed to function internally, one may identify expectations about the outcomes on both individual and collective levels. Individuals should profit from their ac­ tive participation by going through a learning process, clari­ fying their own preferences, and experiencing political effi­ cacy, which in turn should strengthen their role as citizens. But expectations about the results of citizens’ assemblies go beyond the individual level, because their task consists in reaching a collective decision, as a rule one that issues in a specific policy recommendation. The hope here is that a high-quality decision will be forthcoming that has consid­ ered all relevant arguments and reflects the learning curve undergone by the individuals involved(Mercier/Landemore 2012). The goal is not primarily to arrive at a consensus posi­ tion, but to come to a majority decision based on well-in­ formed opinions which, incidentally, may be accompanied by reasoned dissents. Additionally, Niemeyer and Dryzek(2007) put the concept of a meta-consensus on the table as a valid goal for deliberative group processes. A meta-consensus does not imply a unified stance on the recommendations that have emerged from the process of deliberation; rather, it simply refers to agreement about what valid reasons and 8

The discussion in democratic theory legitimate alternatives there are on a given issue, not neces­ sarily how those are to be weighted or rank-ordered. If procedures for recruitment and selection as well as the de­ liberative process itself do justice to those normative princi­ ples, it becomes much more likely that the output of the cit­ izens’ assembly will be recognized as legitimate by the broad­ er public. Moreover, if procedures and arguments are com­ municated transparently, they may also stimulate and inform public debate, thereby reinforcing the democratic culture of discussion. There is another sore point in the theoretical dis­ cussion that also must be considered: the question of what role should be assigned to citizens’ assemblies in the political system. In the final analysis, what is at stake here is the issue of whether—or if the answer is“yes”, how—the recommen­ dations of a citizens’ assembly can acquire authority and, as the case may be, how they should be transformed into deci­ sions binding upon the collectivity. The following systemic considerations are meant to address that question. 2.3 SYSTEMIC REFLECTIONS Recently, there have been some significant trends in the field of mini-publics theory. One of those attempts to parry the criticism that the focus on such publics increasingly distracts attention from the process of political will-formation and dis­ cussion at the level of society as a whole(Chambers et al. 2009). Some critics go so far as to complain that deliberative mini-publics undermine democratic discourse, because to some degree they divert potentially critical discourse from the public sphere. This is(allegedly) particularly true when they are deployed with strategic intentions by state authori­ ties to minimize opposition(Böker 2017; Lafont 2020). Other authors take this warning seriously, but they point out that modern complex societies require a division of labor, even a discursive one, in which deliberative mini-publics potentially have an important role to play(Setälä/Smith 2018). Systemic approaches(classically Habermas as early as 1994) attempt to give this problem due consideration. They seek to clarify how the various political arenas of a democratic system might mesh in ways best suited to fulfill democratic functions(Mansbridge et al. 2012). In that context too, we might well wonder which functions in a democracy could be fulfilled best by deliberation, in addition to other practices (Warren 2017). At the heart of those reflections arises the question of how subsystems are interlinked. That is, what role should deliberative mini-publics such as citizens’ assem­ blies play vis-à-vis representative institutions(parliament and government), the organizations of civil society(parties, NGOs), and the broader public? Basically, theorists tend to favor a loose linkage(Mansbridge et al. 2012). If that were not the case, and the linkage were tighter, it might prevent mini-publics from operating with sufficient independence to realize their inclusive and delib­ erative potential. A closer linkage might even cause mi­ ni-publics to be co-opted by other actors for goals extrane­ ous to them(those of party politics or governmental policy). But a complete disjunction between the two sides would al­ low the possible effects of mini-publics to fizzle out. The question becomes: what does a loose linkage mean con­ cretely for the design of such cooperative relations? The precise configuration of these connections must be re­ garded as dependent upon the functions and authority as­ cribed to citizens’ assemblies and other deliberative mi­ ni-publics. In this respect we can distinguish two fundamen­ tal ideas. Either those arenas(i.e., mini-publics) play a mere­ ly consultative role or else, building on the latter, they will be integrated into the process that generates binding political decisions. Furthermore, they can assume a consultative role either toward the broader public or in respect to political de­ cision-makers. It is possible to do both at once. In the first case, the recommendations prepared by citizens’ assemblies can provide orientation for the citizens, because the latter approve of the reasoning that was communicated to them or because they are confident that these representative cross-sectional groups, standing in for a deliberating public, have come up with the proper positions(Warren/Gastil 2015). In the second case elected decision-makers can take into account the arguments and recommendations of citi­ zens’ assemblies during their own deliberations. Alternative­ ly, they might treat those arguments as reflecting the range of opinions in the public at large. If deliberative mini-publics are drawn more deeply into the political decision-making process, then multiple possibilities emerge. A first variant consists in the tight connection be­ tween the deliberations of a citizens’ assembly and those of a representative decision-making body such as a parliamen­ tary committee. Scholars regard the reciprocal exchange of arguments and the willingness of elected decision-makers to debate the members of the mini-public on the basis of equality as decisive for its success(Hendriks 2016). In a sec­ ond variant, a mini-public is tied to the decision-making procedures of direct democracy. Thus, each set of recom­ mendations made by the citizens’ assemblies of British Co­ lumbia and Ontario, by now regarded as classic instances, was linked to a referendum. In this manner their electoral law reform proposals could be voted on directly. A third variant goes farther still, allotting direct decision-making authority over a specific field to a citizens’ assembly. There are advocates of this position among representatives of di­ rect democracy as well, although fewer than for the op­ tions previously introduced(Setälä/Smith 2018). One prom­ inent example of this variant is the proposal to set up a sec­ ond or third legislative chamber, the members of which would be chosen by lot. Gastil and Wright(2018), who back that position, see a distinct advantage in combining selec­ tion by lot with balloting to determine the composition of different legislative chambers. It would, they claim, coun­ teract the shortcomings of party competition such as the increasing reliance on expensive marketing tools in elec­ tions at the expense of substantive debates. But in that case, in the opinions of Lafont(2015) and Kneip and Merkel (2017), problems of legitimacy could emerge, since arbitrar­ ily selected citizens would lack the duty of accountability to­ ward their fellow citizens. 9

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES The more closely that citizens’ assemblies and other forms of deliberative mini-publics are integrated into binding process­ es of decision-making, the stronger is the tendency for them to compete with other institutions and civil-society actors like parties and interest groups, should the occasion arise. What is more, the incentive increases for political actors to influence citizens’ assemblies for strategic ends or simply to use them on an ad hoc basis to attain pre-established goals. Thus, it is all the more important to give some thought to the responsibilities of the bodies chosen by lot and how they might be incorporated institutionally into representative pro­ cesses of decision-making. To avoid a one-sided capture of the assemblies, we need to reflect on the question of how their substantive agenda might be checked and monitored (Setälä 2017). However these linkages are organized, a crucial condition for the legitimacy of the recommendations approved by the citizens’ assembly is that the broader public should have the maximum opportunity to follow the arguments underlying those proposals. It seems crucial to solicit media attention, organize exchanges with the public, and publicize justifica­ tions for why elected representatives either did or did not take up the suggestions made by citizens’ assemblies. Only when we make sure that in principle the public can under­ stand the reasoning of the citizens’ assemblies can we counter the criticism cited above: that deliberative mi­ ni-publics run the risk of undermining the discourse of the larger society. of different actors and institutions within the system, espe­ cially parties and parliaments, to engage with these new forms of participation. On the other hand, citizens’ assemblies also must be set up such that they can develop their democratizing potential to the optimal extent possible, while their potential weakness­ es are minimized. The preceding theoretical discussion has made possible the formulation of questions that enable us to assess properly the concrete implementation of citizens’ assemblies. They include the following: Do the selection and recruitment of participants succeed in assuring equality of opportunity and descriptive representation? How inclusive and deliberative are the designs of these processes? Is there a constructive division of labor and effective linkage with the institutions of representative democracy, civil society or­ ganizations, and the broader public? 2.4 INTERIM CONCLUSIONS Could the problems plaguing democratic politics analyzed in the introduction be confronted or even solved with the help of citizens’ assemblies? These deliberative forms of demo­ cratic innovation have the potential to attract politically less active people, or those disappointed by traditional institu­ tions, to participate in democratic politics, thereby integrat­ ing diverse societal perspectives. In the best-case scenario they promote a constructive exchange between different positions, contribute to more thoughtful opinion-formation, temper the polarization of society, and thus inform public discourse. They can help narrow the gap between political decision-makers and citizens by playing a constructive role in political decision-making. Under favorable conditions, they can serve to break through encrusted political routines in all these ways. Let those remarks stand as a summary of at least the normative expectations of citizens’ assemblies and other kinds of deliberative mini-publics. However, systemic thinking in democratic theory reminds us that isolated innovations cannot be seen as cure-alls if other expectations and modes of operation within the political system do not also change. In this sense mini-publics should be regarded less as the sole true arenas of democracy and more as instruments by means of which other aspects of the political system and democratic participation can be democ­ ratized(Curato et al. 2017, p. 32). For that to happen we need not only principled openness but also the willingness 10

Political experiences and empirical findings 3 POLITICAL EXPERIENCES AND EMPIRICAL FINDINGS To assess the potential of citizens’ assemblies to bring about democratization, we can draw upon a pool of experiences that by now has become fairly deep. For the most part, the latter have already been taken into consideration in political science research. Besides those scientific analyses, the pres­ ent study also relies on background conversations with se­ lected experts from the social scientific community and the realm of practical politics(cf. the list in the appendix). 3.1 OVERVIEW As previously noted, deliberative mini-publics have ap­ peared in a variety of guises. They may be distinguished from one another by — among other factors — the selection and number of the participants, the duration of their activi­ ty, and the ways in which the results of their work have been utilized. The most prominent variants are the citizens’ assembly, citizen jury, consensus conference, planning cell, and deliberative poll. In 18 of the member states of the Or­ ganization for Economic Cooperation and Development (OECD) as well as in the European Union, they have been carried out on every level—local(52 percent), regional(30 percent), national(15 percent) and international/suprana­ tional(3 percent). Figure 3 Mini-publics in 18 OECD member states and in the European Union national 15% international/supranational 3% regional 30% Quelle: Č esnulaityt ė 2020 b, S. 70. local 52% Most of them were devoted to topics such as urban plan­ ning, health, environment, and infrastructure( Č esnulaityt ė 2020a, p. 70). Over the course of many years, numerous models of representative processes of deliberation have been developed, tested, and implemented all over the world and have been called by different names depending on the context. Given this diversity and the growing signif­ icance of these novel forms of participation, leading ex­ perts from the social-scientific community and the world of practical politics recently have tried to standardize the nomenclature to ensure the quality of such innovations. In this context several contributions should be noted, includ­ ing, among others, the manual“Enabling National Initia­ tives to Take Democracy Beyond Elections” issued in 2019 by the UN Democracy Fund and the newDemocracy Foun­ dation, as well as the study entitled“Deliberative Mi­ ni-Publics Core Design Features”(Curato et al. 2021). Whereas these publications investigate the diverse forms of mini-­publics, the current study concentrates on citizens’ assemblies as one specific subtype. Citizens’ assemblies, which have existed only since 2004, count as the newest of the above-mentioned forms. It also may be true that they represent the most ambitious demo­ cratic format among all of the types of deliberative mi­ ni-publics because they involve a relatively high number of citizens in a temporally extended undertaking designed to engage them politically, cultivate their powers of choice, and help them reach decisions. According to a study of both OECD member states and non-members, typically the work of a citizens’ assembly takes eleven months and on average is carried out by 90 participants( Č esnulaityt ė 2020b, p. 35). What makes this form of participation especially intriguing is the fact that hitherto it has not focused exclusively on lo­ cal or regional projects, but instead on issues relevant to the entire society: ones that touch on more prominent political topics and positions taken by parties. For example, right now citizens’ assemblies increasingly are seen as part of so­ cieties’ overall responses to the fallout from the COVID-19 pandemic. Not long ago the French city of Nantes founded a citizens’ assembly in which 80 local residents were asked to evaluate the effects of the COVID-19 pandemic and pre­ pare recommendations(convention citoyenne 2021). Their report was to be presented to the city’s elected officials who in turn were supposed to take positions on the propos­ als submitted. The state(Land) governments of Baden-Würt­ 11

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES temberg, Thuringia, and Saxony set up citizens’ assemblies on the topic of“corona,” the findings of which were in­ tended to inform the states’ policymaking. And recently, cit­ izens’ assemblies have been grappling more often with cli­ mate issues. There is little risk in predicting that the demand for civic co-participation on this very issue will continue to grow. Citizens’ assemblies have gained added significance as a form of participation, especially in the last few years. Yet, al­ though they have tended more and more to become the popular institutional answer to hot-button political ques­ tions, judging by their significance within the overall political system they still must be regarded as“marginal institutions” despite the growing attention paid to them by the media and social science literature(Setälä/Smith 2018, p. 300). 3.2 A COMPARISON OF PARADIGM CASES Between 2004 and today a steadily growing number of cit­ izens’ assemblies has been in operation. Among the most prominent cases are the following: the citizens’ assemblies set up to consider reform of the electoral laws in British Co­ lumbia(2004), Ontario(2006), and the Netherlands(2006); the citizens’ assembly process devoted to reform of repre­ sentative democracy and party finances in Estonia(2012– 2014); the Irish Convention on the Constitution(2012– 2014) and the Irish Citizens’ Assembly(2016–2018); the as­ semblies on the climate in France(2019–2020), Great Brit­ ain(2020), and Germany(2021); and both of the German citizens’ assemblies on democracy(2019) and on the coun­ try’s role in the world(2021). Whereas the aforementioned examples were of the one-off variety, each involving a unique, now-completed project, the citizens’ dialogue in East Belgium offers the sole case to date of a permanent citizens’ structure at the state level. There, one finds a citi­ zens’ council that is firmly anchored in the parliament and is responsible for convening citizens’ assemblies, chosen by lot, to consider specific subjects. The composition of the body also undergoes constant renewal, as one-third of the members are replaced semiannually(cf. https://www. buergerdialog.be/). The council, composed of 24 members chosen by lot, decides on the topic for each citizens’ dia­ logue. It may determine the topic on its own or else choose from among submissions from parliament or the public. The parliament and government have a year to decide whether to implement the recommendations of the citi­ zens’ assembly. Although they are not obliged to put those suggestions into practice, they must provide express justifi­ cations for a negative decision(Citizens’ Dialogue of East Belgium 2021). It is striking that so many of these nationally and internation­ ally high-profile citizens’ assemblies concentrate on issues as­ sociated with the democratic system itself. They focus par­ ticularly on reform of electoral laws, but also tackle constitu­ tional questions that establish the framework for democratic politics. Furthermore, the planned but still not empaneled Citizens’ Convention on UK Democracy deals with topics that fit into the same category. It is supposed to deal with fundamental questions concerning the structure of the dem­ ocratic system, including a reappraisal of the prerogatives and membership of the House of Lords(The Upper House), a review of the relationship between central and local admin­ istrations, the issue of decentralization in England, and the legal recognition of certain features of constitutional law, in­ cluding individual rights(CCUKD 2020, p. 13). Because of their non-partisan character, citizens’ assemblies look almost tailor-made for the task of putting forward reforms of de­ mocracy. But they also seem like the preferred means to ad­ dress other challenging issues or long-term problems like cli­ mate change that, for structural reasons, political and state actors find notoriously difficult to handle. The spectrum of topics they can take on ranges from highly specific tasks such as the reform of electoral laws to the formulation of general guidelines, as was the case with the citizens’ assem­ bly“Germany’s role in the world.” The citizens’ assemblies referred to above were launched by different actors. Unsurprisingly then, their relationship to po­ litical parties and the institutions of the state has been high­ ly variable. Frequently, a government is the initiator and “sponsor” of a citizens’ assembly, as happened with British Columbia, where the soon-to-be premier, Gordon Campbell, had already promised while in opposition to convene such an assembly should he be elected to the province’s highest of­ fice. Likewise, the French citizens’ assembly on the climate was empaneled by the executive, President Emmanuel Ma­ cron. Still other initiatives originated with civil society and, as the process took its course, gained the support of the parlia­ ment and government. That was the case with the first two German citizens’ assemblies on“democracy” and“Germa­ ny’s role in the world.” The topic of the latter emerged from negotiations between the NGO Mehr Demokratie e. V.(more Democracy) and the leaders of the respective party groups in the federal parliament, the Bundestag(cf. interview with Claudine Nierth). By the same token, the origins of the Irish assemblies can be traced back to the engagement of a group of political scientists who launched a pilot project known as We the Citizens. It was the model for the Irish assemblies that were to follow, which received their commissions from the government and parliament(Farrell/Suiter 2019). Still other projects were conducted primarily as initiatives of civil socie­ ty, as exemplified by the German federal-level Citizens’ Assembly on the Climate. It was initiated and sustained by an association known as BürgerBegehren Klimaschutz e. V.(Cit­ izens’ Referendum Petition on Climate Protection) which was reacting to the fact that the parliamentary groups in the Bun­ destag had failed to include the issue of climate policy in their charge to the previous citizens’ assembly(cf. interview with Nicole Hartmann and Felix Nasser). Although here, too, the input of the parliamentary groups in the Bundestag was sought and incorporated, this citizens’ assembly remained an affair run exclusively by civil society, albeit within a broad al­ liance of supporting civil society organizations. We may con­ sider the Estonian citizens’ assembly process as an outlier in this lineup, since it was founded entirely without prior coor­ dination with political parties; indeed, at bottom it could be understood as a kind of anti-political party operation. It was 12

Political experiences and empirical findings responding to the political crisis occasioned by the revelation of a secret financing scheme that was intended to bypass the legal regulations on political party financing(Karlsson et al. 2015). The process was initiated by the country’s president, Toomas Hendrik Ilves, in cooperation with organizations of civil society, and focused on the electoral system and party finances, among other issues. Against this backdrop, a variety of motivations for instituting and supporting citizens’ assemblies can be identified. As far as political decision-makers are concerned, experimentation itself is sometimes the main attraction: i.e., their interest in testing out the new format to see what opportunities and risks it harbors. But there is a second incentive for state ac­ tors to utilize citizens’ assemblies: namely, the prospect that they might deliver a more subtle picture of public opinion than that derived from the usual ad hoc opinion surveys. Governments, especially, often hope to obtain answers from citizens’ assemblies that will help them manage complicated, hotly contested issues. We can also discern a higher-order motivation here: to regain trust or to secure legitimacy. That was the rationale behind the government of East Belgium’s decision to start a permanent citizens’ dialogue, although in time the entire parliament rallied to support it. In this way, the threat of populist tendencies in particular should be countered by bringing citizens closer to political deci­ sion-making processes again(cf. interview with Anna Quad­ flieg and Anna Stuers). But then again, from the standpoint of civil society the goal may be more fundamental: to pro­ mote civic participation or to put issues on the agenda that established politics has not taken up, at least not enough to suit activists. In the empirical cases we have surveyed, citizens’ assem­ blies usually perform a complementary function. They were established to take on tasks that traditional demo­ cratic institutions and actors alone do not handle very well. There is no question that the processes of citizens’ assem­ blies can engender tense relationships with the actors and routines of representative democracy. This is so because their procedures follow a peculiar logic at odds with the dominant patterns of representative democracy, which re­ ly heavily on influence and competition. It is for that very reason that they can offer support to political representa­ tives who want to pursue projects that transcend the limits of competitive electoral cycles. Citizens’ assemblies also can contribute to obtaining a hearing for civil society per­ spectives in the political process. Nevertheless, one essen­ tial characteristic of a well-run citizens’ assembly is that its outcome should be open. The citizens’ assemblies noted already definitely generated some recommendations that came as a surprise to their respective initiators and observ­ ers, including a relatively complex electoral system(British Columbia), a liberalization of marriage laws(Ireland), and some remarkably radical climate policy measures(Germa­ ny). And that is why it presents a challenge to explain to political decision-makers or activists, why they should en­ gage in a process they cannot control and that might lead to outcomes they would not endorse(interview with An­ drew Blick). 3.3 THE BALANCE SHEET How should the democratic performance of the featured citizens’ assemblies be judged? Below, we will weigh sever­ al factors: aspects of representativeness and external inclu­ siveness; the quality of procedures and internal inclusive­ ness; the effectiveness of outputs; and the financial costs associated with the establishment of those assemblies. REPRESENTATIVENESS AND EXTERNAL INCLUSIVENESS One of the crucial promises of citizens’ assemblies is that they facilitate the inclusive participation of people who, tak­ en as a whole, embody or mirror the demographic charac­ teristics of their society. To begin with, a study of the British Columbia Citizens’ Assembly shows that its relatively strict random selection, which focused only on balancing aspects of gender, regional origins, and age, produced a circle of participants that manifested a certain diversity in respect to other factors such as ethnic background, professional sta­ tus, and education. However, because no representatives from the indigenous community were randomly selected, two additional places later were assigned to members of that group. Still, the selection of members did not achieve a strictly descriptive representation of the province’s popula­ tion. Presumably due to self-selection, the participants were disproportionately white, older, salaried employees or retir­ ees with university educations(Warren/Pearse 2008). James (2008) argues that the organizers should have stratified their random selection so as to take ethnic and socio-eco­ nomic backgrounds into account as well, because those factors might correlate with different attitudes toward the electoral systems examined in the citizens’ assembly. Similar disproportionalities were to be found in Ontario and the Netherlands even though both likewise exhibited a certain diversity. In all three citizens’ assemblies, people with uni­ versity degrees and with high levels of political interest were overrepresented(Fournier et al. 2011, ch. 3). In the Irish cases, in which socio-economic status was built into the stratification of the random selection, the hoped-for descriptive representation evidently turned out quite well (Farrell/Stone 2020). For its part, the evaluation of the fed­ eral-level German citizens’ assembly on democracy revealed that people with higher educational attainments, positive attitudes toward more civic participation and generally high political interest were overrepresented(Geißel et al. 2019). Nevertheless, the proportion of the participants with only a certificate of secondary education was higher than it usual­ ly is in elections(interview with Brigitte Geißel). The spon­ sors of the citizens’ assembly on climate believed that they were facing a similar challenge and saw that there was a need for improvement here as well(interview with Nicole Hartmann). This survey shows that, even though previous citizens’ as­ semblies may not have attained perfectly descriptive repre­ sentativeness, they always have succeeded in getting a di­ verse range of average citizens to participate. So, in principle 13

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES this is a format that makes it possible to reach people who otherwise are not politically active(interview with Brigitte Geißel). That assumption is supported by the results of stud­ ies showing that people who are skeptical about conven­ tional forms of participation, nonetheless welcome its delib­ erative versions(Neblo et al. 2010; Webb 2013). For that matter, existing problems with representation do not ap­ pear insoluble. They can be mitigated by stratification suita­ bly adapted to the circumstances. Therefore, in support of such mitigation efforts, selection procedures for participa­ tion must be devised that target underrepresented and mar­ ginalized groups(cf. chapter 4). Ultimately, we will never be able completely to eliminate the role of interest in politics or in the envisaged topic as factors influencing a person’s deci­ sion on whether to participate. But even here it is not hard to think of measures that would stimulate political interest and attention. Efforts to broaden external inclusion and representativeness must begin even at the assemblies’ agenda-setting stage. The case of the Estonian citizens’ assembly exemplifies that point. Using the crowdsourcing method, 2,000 proposals and 4,000 commentaries were collected in Estonia(Karlsson et al., p. 6). With the assistance of representatives of civil so­ ciety organizations as well as a group of volunteer experts in political science and law, 18 proposals subsequently were identified that served as the basis for the work of the citi­ zens’ assembly. A closer analysis, however, revealed that the majority of those who took part in the crowdsourcing were already politically active males(Jonsson 2015). And in the case of the regional conferences that were to set the agen­ da for the citizens’ assembly on democracy, Geißel et al. (2019) found that systematic biases in respect to education, political interest and political attitudes came to light that were more strongly marked even than those in the citizens’ assembly itself. Thus, even when the citizens’ assembly represents a mirror image of society, that still is no guarantee that the substan­ tive deficits in representation will be completely erased. The examples cited here make it clear that the problem of defi­ cits in representation crops up again when the citizens’ as­ sembly deliberates about previously framed proposals, in the development of which a majority of privileged social groups were involved. It often times turns out that the de­ sire to participate is less evident among those fed up with politics“than it is among those who are politically interest­ ed anyway, for whom voting every four years is too little” (Kübler et al. 2021, p. 47). This is a point that also must be taken into account in the design of the citizens’ assembly process. THE QUALITY OF PROCEDURES AND INTERNAL INCLUSIVENESS Many of the aforementioned citizens’ assemblies were eval­ uated in light of the quality of their internal procedures. As a rule, questionnaires submitted by the participants and ob­ servations by external scholars serve as the basis for such evaluations. Over 90 percent of the participants in the citi­ zens’ assemblies of British Columbia and Ontario agreed with the statement that all members had an equal opportu­ nity to express their views(in the Netherlands that number was somewhat lower, at 77.7 percent). Furthermore, in those cases mutual respect was given high marks. The same is true of satisfaction with the ultimate decision(Fournier et al. 2011, p. 46). Similarly, the participants in the Irish Constitutional Convention and Citizens’ Assembly said they were highly satisfied with the process. The questionnaires provid­ ed little support for the idea that some members felt domi­ nated by others(Farrell/Stone 2020, p. 238f.). And in the case of the German Citizens’ Assembly on Democracy, both the members of that body and the outside evaluators con­ cluded that the informational and deliberative procedures were of high quality, the facilitation balanced and fair, and the discussions open and respectful(Geißel et al. 2019). In an interview with the authors, Antoine Vergne reported similar perceptions from the point of view of the organizers of the French climate assembly. According to his testimony, procedural facilitation and normative framing enabled the members of the French assembly to develop a high mutual regard for egalitarian and inclusive participation. Moreover, he noted that they were put in a position to see for them­ selves and decide what additional information they still needed. These findings prove that it is possible to imple­ ment the goal of an inclusive citizens’ deliberation of high epistemic quality. Nevertheless, challenges remain in respect to internal inclusion in the process of deliberation that have come to light in both research and practice(cf. section 3.4). THE EFFECTIVENESS OF OUTPUTS How effective are citizens’ assemblies? In the cases treated here the participating citizens managed to submit elaborate proposals for the problem addressed in each case. But were their recommendations considered politically or even put in­ to practice? The Canadian citizens’ assemblies presented proposals for a reform of the electoral system that subse­ quently were submitted to all eligible voters in referendum form and voted on. Although the British Columbia proposal did succeed in winning an absolute majority of the votes counted, it barely failed to get past the previously mandated threshold of 60 percent. The Ontario proposition, which carried 37 percent of the vote, failed to gain a majority, while the minimalist reform proposals in the case of the Netherlands were not implemented by political deci­ sion-makers in the wake of the fall of the governing coali­ tion there(Fournier et al. 2011, p. 49f.). Fournier and his col­ leagues(2011, ch. 8) explained the failures in the two Cana­ dian cases by citing the weak commitment of elites as well as the lack of media attention in the runup to the referen­ dums. In the case of the Netherlands, they attribute failure to a pervasive lack of interest in electoral reform on the part of the political class. The Irish citizens’ assemblies proved to be more influential in terms of policy outputs. Among other accomplishments, they led to successful referenda on the legal status of same14

Political experiences and empirical findings sex marriage and abortion. Those two issues, which long had divided Irish society, were decided and put to rest by the deliberative treatment of the topics in society and referen­ dums. Based on such experiences as these, observers and practitioners draw the conclusion that the more fully citizens’ assemblies manage to involve political decision-makers in their launch and organization, the more likely it is that they will be successful(Farrell/Suiter 2019, p. 31f.; Antoine Vergne in an interview). Accordingly, the initiators of the federal-lev­ el German citizens’ assemblies made every effort to gain the support of party delegations in the Bundestag and, with their aid, to define the scope of the issues to be addressed. In the process that followed, the relevant parliamentary com­ mittees were able to refer back to those conversations(inter­ view with Claudine Nierth). And in East Belgium the govern­ ment that initiated the citizens’ assemblies, along with rele­ vant scholarly experts, worked hard and effectively to win the support of all parliamentary parties so that lingering skepticism could be eliminated and, in the end, a permanent citizens’ dialogue could be approved unanimously and put into effect through a decree supported by the entire parlia­ ment(interview with Anna Quadflieg). Thus, it seems to be a harbinger of success for citizens’ as­ semblies when the initiators cooperate with parties and elected decision-makers-most of all when issues are at stake that affect the latter directly such as the basic rules of the game in a democracy. Yet cooperation is no guarantee of success. Although at the time of the French citizens’ assem­ bly on climate, President Macron promised to give legal force to the assembly’s recommendations through executive or­ ders, parliamentary approval, or even a plebiscite, to this very day the implementation of the proposed measures has stalled. 4 Nor is the involvement of all the relevant actors al­ ways achieved. Although the Citizens’ Convention on UK Democracy was able to win over high-ranking MPs from differ­ ent parties as political sponsors, its installation has dragged on slowly because the Conservative government hesitates to support the process(interview with Andrew Blick). The ex­ ample of the German Citizens’ Assembly on Climate proves at least that an ambitious, high-quality participatory proce­ dure can be put into effect even without support from the political ranks. Nevertheless, in the estimation of Felix Nass­ er, one of the project’s coordinators, that outcome was pos­ sible only by virtue of an enormous effort, plus the coopera­ tion of various actors from civil society(Felix Nasser inter­ view). It remains to be seen whether the recommendations will have any influence on the policy-making process. In Estonia’s citizens’ assembly process, prospects for accept­ ance of the recommendations looked rather dim, because it criticized the behavior of the parties represented in parlia­ ment, which obviously did not make it any easier to enlist their cooperation. Nevertheless, the procedure led to some notable triumphs. Of 15 proposals made, three ultimately were accepted by the parliament, while the parliament and government considered parts of another four(Rahvakogu 2014). Aside from the reforms already mentioned, the citi­ zens’ assembly apparently had an emancipatory effect. It re­ inforced the citizenry’s perception that it has the right, the opportunity, and the means to be included in political dis­ cussions. As some observers see it, this has contributed to making parliamentary deputies more willing to support the involvement of the public and similar initiatives(Centre for Public Impact 2019). If citizens’ assembly proceedings man­ age to generate high-visibility legitimacy, political deci­ sion-makers cannot afford to ignore the normative pressure that emanates from them. THE FINANCIAL COSTS OF CITIZENS’ ASSEMBLY The costs of a citizens’ assembly depend on a variety of fac­ tors such as the number of participants and meetings, the type of meetings(online or in person), and the institutional structure and connections. The average costs for a citizens’ assembly, judging by their conduct in the European Union, Australia, Canada, Great Britain, and the USA, amount to 1.8 million euros( Č esnulaityt ė 2020a, p. 78). The price tag must be seen in relations to the potential services that a cit­ izens’ assembly can render to society. Into that sum must be factored the service of cost-avoidance: e.g., of the costs that would have been incurred by ignoring a given topic or not getting citizens involved. If citizens’ assemblies manage to contribute to the democratization of democracy and help find legitimate solutions to virulent political problems, they should be regarded as reasonable investments for a demo­ cratic polity to make. However, the financing question is of paramount importance in respect to the impartiality of citizens’ assemblies. Thus, Neblo claims(rightly) that the“extension of use of mini-pub­ lics creates incentives for interest groups to try to manipulate them”(2015, p. 81). To prevent interest groups from hiding behind citizens’ assemblies to advance particularistic inter­ ests, transparent financing over the long term is necessary. Thus, for example, Nanz and Leggewie(2018) suggest that citizens’ assemblies should be financed by a foundation cre­ ated for this express purpose, while Geißel and Jung(2019) recommend that they be funded from the state’s budget. It is essential that citizens’ assemblies be independent as they engage in making decisions. A foundation bankrolled pri­ marily by the state but publicly controlled—perhaps on the model of public broadcasting councils—actually might offer an opportunity to continue financing deliberative forums in transparent and sustainable ways, even if they should be­ come more widespread throughout society. 3.4 QUESTIONS AND CHALLENGES 4 Ten percent of the recommendations were approved unchanged by parliament and executive directives, and 37 percent were accepted in modified versions, while 53 percent were rejected(Courant 2021). The experiences reviewed here suggest that, on the whole, citizens’ assemblies have performed in highly promising, democratic ways. But needless to say, that does not mean 15

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES that daunting challenges do not remain on the way toward the optimal practical implementation of their democratizing potential. As Claudine Nierth put it in an interview, citizens’ assemblies are“learning models.” Building on evaluations and reflection, they may be developed further. We can iden­ tify a variety of“construction sites” where such advances could be made. The first cluster of such challenges revolves around the pre­ viously discussed matter of external and internal inclusion. By now, the question of how to make the most representa­ tive possible random selection work well has been ad­ dressed not only in theory but also in practice, and in sophis­ ticated ways. But the issue of internal inclusion — how to ensure the participation on an equal footing of all the citi­ zens who were selected and the effective integration of their diverse perspectives — also has arisen as a central con­ cern in both social scientific discussions and the interviews conducted for this study. Here, the crucial role of facilitators, above all the need for them to be professional and neutral, has been identified as the key to both internal integration and the guarantee of a high-quality deliberative process(in­ terviews with Brigitte Geißel, Felix Nasser, Claudine Nierth, Anna Stuers, Iain Walker). In the estimation of our interview partners, systematic training might be necessary to ensure the requisite level of professionalism among facilitators. In that same context, interviewees point to the crucial signif­ icance of the experts called upon to provide information to participants. For instance, on the issue of climate policy, sci­ entific communication may prove to be a potential stum­ bling-block unless some way is found to communicate re­ search findings even to people without higher education backgrounds. But at the same time, it is important to avoid a situation in which individual experts, solely on account of their rhetorical skills, exercise a more pervasive influence on the deliberations than do others(as Claudine Nierth noted in an interview). In his interview Iain Walker reported that one constitutive element in the mini-publics that his organi­ zation sponsors is the explicit teaching of critical thinking to participants by professional facilitators. A subject that has become especially topical in the wake of the coronavirus crisis is the role of digital communications formats in citizens’ assemblies. In light of their most recent experiences, the interviewees we consulted for this study see both advantages and disadvantages. One advantage they emphasize is that digital formats can enable less-mo­ bile individuals or those more tied to specific locations due to family obligations to take part in citizens’ assemblies. Fur­ thermore, digital platforms allow topics to be collected in the aftermath of agenda-setting; they permit information to be stored and retrieved, as well as(where appropriate) the written comments showing the positions taken by individu­ als that otherwise might get lost in the shuffle of verbal communications(So argues Claudine Nierth in the inter­ view). On the other hand, there is unanimous agreement that exclusively virtual encounters represent a“hurdle for in­ terpersonal exchanges,” as Felix Nasser put it in the inter­ view. Interview partners such as Iain Walker and Andrew Blick stressed that so far the digital format is not an ade­ quate replacement for a face-to-face setting with its oppor­ tunities for informal communication. The group cohesion that is conducive to a fruitful process of deliberation cannot be generated digitally in the same way it can through in-per­ son contact. There is a consensus among the interviewees that a hybrid format combining, for example, digital pres­ entations by experts with analog phases of deliberation and decision-making could make a lot of sense. One of the major“construction sites” is the question of how citizens’ assemblies might be institutionalized and their rela­ tionship to elected decision-makers clarified. Our conversa­ tion partners think it is important to clarify that relationship prior to beginning the citizens’ assembly process, to make sure that expectations will be managed properly. Once it be­ comes clearer what is to happen to the assembly’s recom­ mendations and in which arenas of political decision-mak­ ing, realistic expectations then can be encouraged and false hopes avoided. When expectations and actual reactions drift too far apart, not only might that widen the chasm be­ tween citizens and political elites, but it also could discredit democratic innovations well into the future(Hanson 2018). In this sense, Anna Stuers, the permanent secretary of the Citizens’ Dialogue in East Belgium, believes that the distribu­ tion of tasks and roles between the Citizens’ Dialogue and the parliament is strictly complementary(interview). Since the decisions of the citizens’ assembly are merely recom­ mendations without binding force, the permanent citizens’ dialogue should also not be regarded as the classic second chamber of parliament. Aside from the question of how the recommendations sub­ mitted by a citizens’ assembly will be received by political decision-makers, it is also interesting to ask what participa­ tion“does to” the citizens(cf. interviews with Brigitte Geißel, Claudine Nierth, Anna Quadflieg). Will they undergo a lasting political socialization that will influence their future participatory behavior and consequently also the broader public? That would be a major win for democracy, especial­ ly if it reached people who otherwise kept their distance from political participation. In an interview with the authors, Andrew Blick points out that the efficacy of citizens’ assem­ blies should not be measured by their direct influence on political decisions alone, but also by the change in political culture that new forms of participation may entail. Still, an enduring cultural transformation can succeed only if citi­ zens’ assemblies really do exert an influence over policy de­ velopment. Finally, one long-term question that we should keep an eye on and continually evaluate concerns the character of the recommendations formulated by citizens’ assemblies. For example, it would be worth asking whether citizens’ assem­ blies’ recommendations have any systematic tendency to­ ward more or less egalitarian solutions. Critics worry that new, deliberative forms of participation serve less as the means of emancipating citizens than as ways to contribute to the legitimation of power relations and inequalities(cf. Walker et al. 2015). Therefore, it is crucial always to examine 16

whether and to what extent procedures really have been rendered neutral toward outcomes. Admittedly, it is hard to think about democratizing democracy without having basic trust in the political judgment of average citizens. But still, the faculty of judgment must be backed up by an optimal design of the process. Political experiences and empirical findings 17

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES 4 THE DESIGN OF CITIZENS’ ASSEMBLIES Thus far we have summarized the expectations about citi­ zens’ assemblies by relying both on democratic theory and on the lessons derived from previous experience with them. Now it is time to sift the relevant results of the research lit­ erature and discuss the challenges posed by and the possi­ ble solutions for the design issues that are so vital to en­ hancing the democratizing effects of citizens’ assemblies. ing, which implies certain answers to the following ques­ tions: – When and in which cases should citizens’ assemblies be convened? – Who is able and entitled to initiate a citizens’ assem­ bly? 4.1 BASIC PRINCIPLES: FLEXIBILITY IN DESIGN, BINDING AND TRANSPARENT GOALS – Are the assemblies always to be set up on an ad hoc basis, or is an option for permanent status conceiva­ ble? As keys to the organization of citizens’ assemblies and their adequate embedding in the political system, we propose an emphasis on flexibility in design while giving due considera­ tion to transparency and bindingness of its specific goals. We are guided by the conviction that all citizens’ assemblies share a common ground plan, the concrete elaboration of which depends on the way the problem is posed and how the goal associated with it is carried to fruition(Farrell et al. 2019). That does not imply an excessively laissez-faire treat­ ment of formats. On the contrary: the design of the assem­ blies must do justice to the basic aspirations to inclusive rep­ resentativeness, deliberative quality, and systemic efficacy anchored in democratic theory. The flexibility in design ap­ plied to the establishment of a citizens’ assembly ought to go hand in hand with concrete transparency of goals and secu­ rity of expectations. Only when the function, task, and remit of the concrete citizens’ assembly are clear and binding can the appropriate decisions about its further elaboration be made and its internal and external communication processes acquire the power to generate legitimacy. For example, a cit­ izens’ assembly charged with developing a general problem orientation on a new political issue can and must have a dif­ ferent structure and systemic embedding than one that is supposed to prepare decision proposals meant to solve a concrete social conflict case. However, it must always be clear and binding what is to be done with the recommenda­ tions. Otherwise, the whole project runs the risk either of be­ coming a pawn of strategic external actors or simply of gain­ ing no traction at all and thus evoking frustration on the part of both the participants and the public. – Who determines the assembly’s agenda and what will the range of topics be? – What kind of lottery scheme will be used to select and recruit the participants? – How must the process be designed so that it satisfies the criteria of inclusivity and epistemic quality? – What type of budget, whether of time or money, is needed to fulfill the envisaged task? – Who will support and organize the citizens’ assembly? – How can the broader public be involved in the pro­ cess? – Should the citizens’ assembly just produce a set of opinions that will be incorporated into the processes of will-formation and decision-making, much like oth­ er kinds of information? Or should it formulate an au­ thoritative recommendation that would become the starting point for a decision that eventually will re­ quire ratification either through plebiscite or by parlia­ ment? – Should the citizens’ assembly even be granted broad decision-making authority over a specific range of is­ sues, as sometimes has been the case, for instance, with the second chamber of legislatures? The task of defining the objectives of function, task, and re­ mit largely is a matter for democratic political decision-mak­ – How will such an assembly be embedded in the insti­ tutions and procedure of representative democracy? 18

The design of citizens’ assemblies Under suitable conditions, all of these optional goals and item categories may be legitimate and consistent with the ways in which citizens’ assemblies work. Some of them require similar design choices, others different. It does not seem plausible to create blueprints for every one of these cases. While the elab­ oration may vary in each case, the crucial point is that the democratizing potential of citizens’ assemblies should come to the fore and their deployment should be perceived as legiti­ mate. Maintaining the quality standards dictated by democrat­ ic theory — representative inclusion, procedural quality, and systemic efficacy — is one indispensable condition that must be met if each project is to redeem its promises. What is at stake here is a judicious weighing and balancing of expecta­ tions and design decisions. Thus, little by little we can benefit from positive experiences and learn from mistakes. Together they can continue to enhance the future developmental pros­ pects of this innovative form within the democratic system. As long as the goal and function of a citizens’ assembly are clear, there are basically no theoretical limitations with regard to possible topics and materials — apart from the desidera­ tum that they should address relevant political questions. Pre­ cisely because goals and tasks vary so much, they have varia­ ble design implications. One obvious but far from trivial con­ sequence of the way in which goals and tasks are defined is the requirement that there should be an adequate budget and time frame for the assemblies’ work. The funding for logistics, organizational costs and, where appropriate, ex­ pense allowances should be tailored to the magnitude of the project. The same logic holds true for the allotted time frame. The broader and/or more complex is the definition of the tasks at hand and the greater is the need for information pro­ cessing, consultation, and deliberation, the more time should be reserved for the process(Farrell et al. 2019). To cite an ex­ ample, the first citizens’ assembly on electoral reform in Brit­ ish Columbia met for nearly a year before presenting its well-vetted recommendation. Also, to be able to clarify these budget issues in advance, the assemblies will need a clear outline of their tasks and objectives. The time frame issue, however, not only touches on the relationship between the work that has to be done and the time budgeted for it, but al­ so the timing of and synchronization with other proceedings of the representative political system — aspects always also bound up with matters of power and influence(Schäfer/ Merkel 2021). Thus, whether a citizens’ assembly presents its results in the middle of a legislative period or at the end could have some bearing on whether it is noticed by other political actors and the general public and how high its profile is. In some cases, a shift in which party or parties hold a majority in parliament after elections can cause the chances of a propos­ al to dwindle. That, in turn, hinges on who initiated the citi­ zens’ assembly in the first place and thus who has an interest in making sure that its results do not fade into oblivion. 4.2 AND AGENDA-SETTING In this context, three intertwined issues are relevant: Who should be entitled to initiate or convene a citizens’ assem­ bly? Who determines the agenda of this citizens’ assembly? Does the citizens’ assembly take place on an ad hoc basis, as a one-off event, or does it form part of a larger institutional framework? As long as the organization of a citizens’ assembly is not regulated by(constitutional) law, anyone is free to start one. Its subsequent success or failure will depend on how much political support the founders can gain and what resourc­ es — above all how much political influence, funding, and symbolic capital — they have in reserve. Many mini-publics are started by governments, administrations, and public of­ ficials, usually with the goal of finding solutions to serious political conflicts or to acquire a more nuanced picture of public opinion. No basic objections need to be made against this top-down variant of citizens’ assembly start-ups as long as the independence of the proceedings is respected and a co-opting of the participants due to an excessively close linkage to the executive can be avoided. Their democratiz­ ing potential, however, depends strongly on the openness of the initiators and their respect for the citizens’ assembly and its recommendations. If the citizens’ assembly does have rather intimate ties to the political decision-makers who launched it, this will in­ crease the likelihood that its results will be considered seri­ ously and thus that it will succeed. But at the same time there is a danger that it will be misused as a tool to gener­ ate legitimacy(Caluwaerts/Reuchamps 2016). Even if that danger is avoided, it can easily happen that, under these circumstances, the citizens’ assembly will be used as a more upscale kind of focus group or that the decision-makers who initiated it — whether from government, administra­ tion, or parliament —  will single out only those aspects of the recommendations that they like(Setälä 2017; Font et al. 2018). Therefore, once the citizens’ assembly has been con­ vened, it should have a chance to help shape its own agen­ da. A rigid dictation of its thematic tasks that powerfully prefigures the direction of the outcome, would considera­ bly narrow the democratic autonomy of the body and thus limit its democratizing potential. At the other end of the spectrum, there are also prominent examples of the bottom-up approach to starting citizens’ assemblies such as the“We the Citizens” group that was or­ ganized by the Irish Political Studies Association and fur­ nished the model for the later Constitutional Convention and Citizens’ Assembly, both established by the government and parliament(Farrell/Suiter 2019). Also, the three previous citizens’ assemblies held at the federal level in Germany be­ gan with initiatives from civil society. Whereas the first two succeeded in gaining the support of the Bundestag, the cit­ izens’ assembly on climate was started and sustained exclu­ sively by civil society. Naturally, initiatives that originate en­ tirely from civil society have a greater potential to upset the existing routines of the political system by opening up new possibilities for participation, putting new issues on the agenda, and introducing previously unheard voices into the political process. Then there is a tendency for the“disrup­ tive” aspect to culminate in a trade-off relationship with the 19

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES political influence that the results could have on political de­ cision-making(Caluwaerts/Reuchamps 2016). Not least be­ cause they want to enhance the prospects that their recom­ mendations will be considered seriously, the initiators of the assemblies often will enter into negotiations with political decision-makers in regard to the citizens’ assembly’s agen­ da. Here, too, what must happen is that the assembly itself should enjoy a bit of maneuvering room to modify its origi­ nal agenda so as not unduly to constrain the open-ended­ ness of the proceedings. The process itself must go forward independently of the interests and preferences of the initia­ tors and supporters. One great advantage of the ad hoc citizens’ assemblies is that this participatory format can be tested; consequently, a shift in perspective concerning the future development of civic participation may occur among the participants and the general public. But in the long run citizens’ assemblies will not be able to develop their fullest potential for democ­ ratization and exercise a lasting influence on processes of political decision-making until they are embedded compati­ bly in existing legislative and executive procedure, ideally so as to complement the latter while expanding participation (Setälä 2017). The additional integration of direct, delibera­ tive participation into the system of representative democra­ cy does not mean that representative institutions will be weakened, but that they will be bolstered through more in­ formation, reflection, and the consent of the citizenry. So far, three ways have been tested to institutionalize delibera­ tive forms of participation; in each case the concrete deci­ sion about which approach would be best has depended on the local socio-political context(Escobar/Elstub 2017). First: there may be a permanent public structure for citizens’ deliberations, responsible for the topics, agenda, and super­ vision of the ad hoc forms of deliberation like citizens’ assem­ blies as well as for contributing ideas on a specific public is­ sue(Chwalisz 2020, p. 127). Examples of this way of institu­ tionalizing forms of participation may be found in the previ­ ously mentioned East Belgian model, Toronto’s Planning Re­ view Panel, the Citizens’ Initiative Review in the state of Ore­ gon(USA), and the short-lived Observatorio de la Ciudad in Madrid. However, none of those cases unfolded at the na­ tion-state level. This kind of institutionalization comes closest to what some authors have advocated under the name of a “second”(Gastil/Wright 2018) or“third chamber”(Van Rey­ brouck 2016). In this vein Kübler, Leggewie, and Nanz also propose to establish a permanent citizens’ assembly that would firmly institutionalize a clear, obligation-based rela­ tionship to the democratic institutions at the level of the state. Nanz and Leggewie(2018) call it“the consultative branch,” as though it were a third chamber. Second: Public authorities may be obliged to organize a de­ liberative proceeding under certain circumstances. So far, this path to institutionalization has been tied to specific top­ ics. For example, the 2017 Mongolian law on deliberative polling provides that deliberative polls must be organized for constitutional amendments(Stanford News Service 2017). A French law institutionalized the obligation to or­ ganize public debates and consultations in order to change laws that touch on bioethical matters(Chwalisz 2020, p. 123). With that in mind one could make it mandatory to al­ low for citizens’ input whenever amendments to the consti­ tution were being considered. By that stipulation one would bolster the citizens’ power to act as well as deepening their trust in the democratic system and its concrete functioning. Third: Citizens may also be given the option of demanding a deliberative process on a certain topic provided that enough signatures can be gathered. For example, this is the case in the Polish cities of Gda ń sk, Kraków, Lublin, and Pozna ń as well as in the Austrian state of Vorarlberg(Chwalisz 2020, p. 124). In addition to this form of citizens’ initiative, Setälä (2017) suggests that a parliamentary minority ought to be given the formal right to convoke a citizens’ assembly. That might also strengthen the role of the opposition in parlia­ ment and instill a new deliberative dynamism into parlia­ mentary proceedings, which has been lacking until now due to the principles of majority rule and proportional rep­ resentation(Schäfer 2017a). 4.3 AND RECRUITMENT OF PARTICIPANTS As mentioned previously, in addition to the inclusive and epistemic quality of the process itself, the selection of the participants constitutes another important pillar of legitima­ tion for every citizens’ assembly. Random selection by lot makes possible an egalitarian and inclusive procedure. It is intended to maximize equality of opportunity and produce a participant structure that embodies a representative cross-section of the population. Nevertheless, as a general rule that outcome cannot be guaranteed unless the random selection procedure is leavened by stratification. The draw­ ing of lots is carried out so as to take into account relevant categories of the social structure like gender, education, and region(Smith 2009, pp. 80–83). By contrast, with a large population(e.g., at the national level) a pure random selec­ tion would run the risk of failing to achieve the desired di­ versity of participants for the questions to be decided. Be­ sides, a lack of basic information about the population could lead to biases if, for instance, voter registries intended to serve as stand-ins for the population were incomplete or if there were other reasons(e.g., homelessness) why certain strata of the population could not be reached through the invitation procedure. Therefore, depending on objectives and topics, the criteria governing the stratification of the lottery must be specified in order to attain the requisite representativeness(Steel et al. 2020). These criteria ought to be stated and justified explicit­ ly. To do so is all the more important if they transcend the usual, very general social-structural factors. Characteristics such as gender, regional origin, and education may always appear relevant, but for specific topics additional categories may seem worth taking into account: party affiliation when the issue is a possible reform of electoral laws; income differ­ ences when tax policy is up for discussion; or sexual orienta­ 20

The design of citizens’ assemblies tion if the controversies involve family or social policies. Ini­ tially, those characteristics might have to be identified within a larger random sample. On this basis, a stratified lottery could be devised. The basic principle in any reflection upon relevant criteria is that groups significantly affected should be included — especially those whose perspectives and inter­ ests often get marginalized in the regular political process, whether because they tend not to participate or because their power to negotiate is comparatively slight. In light of these challenges, we recognize that there may be a trade-off between the ideal of a pure random selection that offers everyone an equal chance to be chosen by lot and the creation of a kind of diversity suited to the particular is­ sue to be addressed(Steel et al. 2020). This point emerges clearly when we consider the problem of self-selection. Par­ ticipation in a citizens’ assembly is voluntary. Therefore, it is entirely possible that a significant slice of those who have been invited will choose — for whatever reason — not to take part. As long as those reasons are of a purely individual na­ ture, we are dealing with a practical problem of follow-up re­ cruiting. But if the refusal to participate correlates with so­ cial-structural characteristics like income, education, or sta­ tus, then we are confronted by a bias that is troublesome for democratic theory. Since empirical studies prove that just such a correlation exists with participatory behavior(Farrell/ Stone 2020, p. 235; Schäfer 2010), the organizers of citizens’ assemblies will have to take active measures to deal with the problem so that the democratizing potential of the assem­ blies is not jeopardized. The stronger the role of self-selec­ tion is, the more indispensable is a targeted, stratified ran­ dom sampling that may even allocate places by a quota sys­ tem should the occasion arise. In addition, it may be necessary to make extra efforts to re­ cruit participants from certain groups that suffer from a struc­ tural deficit of participation. Thus, those in charge of selec­ tion can locate, seek out, and solicit the involvement of peo­ ple from lower-income groups or those with less formal edu­ cation who are underrepresented in many forms of participa­ tion. In this context it would also help to offer them an appro­ priate allowance. As a rule, in previous citizens’ assemblies costs for travel and lodging were covered, whereas allowanc­ es were handled in a variety of ways(Harris 2019, p. 49). To lower the motivational hurdles for participation, the payment of“daily allowances” of this type can be a useful tool. In order to induce socially marginalized and politically disillu­ sioned people to join in a citizens’ assembly, it is vital to let them know in clear language what its goals are and that these are binding. This is the case because only when they entertain the hope of being listened to and taken seriously during the process will they gain motivation and self-confi­ dence in the sense of“political efficacy”(Jacquete 2017). Moreover, anxiety about interpersonal contacts in the delib­ erative forum can be reduced through efforts at public en­ lightenment: i.e., making explicit the openness of subject matter and the inclusiveness of procedural rules. Additional­ ly, when specific topics such as reform of social welfare pay­ ments are up for discussion, members of affected, under­ privileged groups can be offered a moderated forum in the spirit of the“enclave deliberation” described above. There, they can initially exchange impressions and ideas with one another before they engage in the deliberations of the more diverse citizens’ assembly. 4.4 PROCESS STRUCTURE Citizens’ assemblies should enable average citizens to work through political issues properly and constructively, but also in an inclusive, fair manner. To guarantee that, the process must play out within a clear, binding, well-thought-out structure. The basic structure developed in the by-now classic Canadi­ an citizens’ assemblies provides a worthwhile model that di­ vides the process into four sequential phases, each of which builds on its predecessor(Fournier et al. 2011). On this basis, the Canadian template can be elaborated in the following scheme: The process begins with a learning phase. The complexity of the topics and the participants’ consequent need for information determine how long it will last. Ac­ cordingly, this phase can stretch out over several weeks(or weekends), stimulating both individual and joint learning via mutual exchanges that will enable the participants to appre­ ciate the complexity of the issues under consideration(Esco­ bar/Elstub 2017). The learning materials should be written in an easy-to-understand language and be accessible to every participant. Likewise, when it comes to selecting experts (who may be either scholars/scientists or stakeholders), the principle of diversity rules. The experts called upon to testi­ fy should either be neutral or, on contested matters, they should represent the opposing positions to be taken in each case. To be consistent with the democratic self-image of a citizens’ assembly, such presentations should not impart knowledge in a hierarchical fashion. Rather, they should fea­ ture constructive exchanges that also take into account the “local” expertise offered by the respective participants (Roberts et al. 2020). In this phase of information transfer and learning, the foundations for the following debates and deliberations will be laid. So, it is all the more important to avoid one-sided, partisan information, whether intentional or not. For that reason, supervisory bodies monitoring the process and even the participants themselves should have the option of scrutinizing the choice of experts and, if need be, of supplementing it. Next comes the consultation phase, in which the partici­ pants hold open hearings when and where appropriate in order to gather information and opinions from the general public germane to the issue at hand(Escobar/Elstub 2017). In the course of these two initial phases, the participants should gain access to a plethora of relevant information and expert knowledge. And they should have the option of ask­ ing for additional information above and beyond what has been given to them. Furthermore, all of the experts involved should be treated as equals and have comparable opportu­ nities to present their perspectives to the participants (Chwalisz/ Č esnulaityt ė 2020, p. 102). 21

Figure 4 Process structure of citizens‘ assemblies FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES i Learning Input from independent experts and/or diverse stakeholders ?? ? ? ? Consultation Exchange with broader public and collection of further perspectives Deliberation Results and experiences of the previous phases are discussed and evaluated Final Outcome Decision on the final result, if necessary majority principle with justified minority votes In the third phase the actual process of deliberation takes place. Here the participants, mostly meeting face-to-face in small groups, will have a chance to discuss, evaluate, and ponder the results, experiences, and arguments of the pre­ vious phases(Escobar/Elstub 2017). The decisive point is that they should have enough time for this stage, but also a firm deadline to finish. That is the case because phase three serves the purpose of developing a set of common positions on the concrete political issue before them. As the discus­ sions continue, socialization effects may and will start to happen of the kind that Tocqueville forcefully advocated in pointing to“his” voluntary associations as schools of de­ mocracy. The participants listen, get socialized into empa­ thetic understanding, and exchange arguments. In the wake of those experiences, they find opportunities to figure out— or rethink and in some instances to correct and revise—their own positions. During this segment of the process, in the Canadian and Irish citizens’ assemblies the participants met on average for 18.8 days, and the average elapsed time from the first to the last meeting was 47 weeks( Č esnulai­ tyt ė 2020b). The whole process concludes with the fourth phase, in which the final outcome is decided. Here, there are differ­ ing options for the decision rule, which can also influence the participation of groups underrepresented in conven­ tional forms of participation and representation. A few au­ thors argue that the goal of the process of deliberation should be to reach a consensus or at least an agreement that rests on several differing but reasonable grounds ac­ ceptable to everyone(Eriksen 2009, p. 51). The advocates of consensus argue that the requirement for unanimity has several benefits. It encourages equal respect, allows all par­ ticipants to be recognized as important, and enables every voice and perspective to be accorded proper consideration (Mendelberg et al. 2014). According to this line of argu­ ment, only such a consensus makes room for the full inclu­ sion of marginalized perspectives, underprivileged groups, and/or minorities. The trouble is that an accord based on reasons that com­ mand full agreement or one that would be acceptable to everyone may be unattainable in many cases. Especially when it comes to hot-button issues or socially polarized conflicts, it appears that a consensus, at least in societies with a plural value structure, is neither possible nor abso­ lutely worth having. In such cases, according to Habermas (1996) and Mansbridge et al.(2010), the clarification of con­ flict must itself become a topic of debate, with compromise as the outcome participants strive to attain. Deveaux(2018, p. 160) emphasizes that this pattern often emerges in cases of profound moral conflicts. She adds that participants with different standpoints or world-views must accept a fair compromise in order to arrive at an understanding or settle­ ment not based on coercion or majority decision. According to Gutmann and Thompson(2018, p. 908), readiness to seek compromise, in which everyone gives up something in order to attain the best possible solution for all, is a proce­ dural option for overcoming stubborn partisan blockades. Incidentally, compromises are deeply inscribed in the gram­ mar of democracy. As such, they can exert a positive, exem­ plary influence on broader circles of the population. 22

The design of citizens’ assemblies When serious differences of opinion prevail, an apparent consensus can conceal the fact that those less skilled in ar­ gument will sometimes be pressured to submit to the will of the majority(ibid., p. 907) or that existing conflicts will be si­ lenced by demands for rationality(Schäfer/Merkel 2020, p. 462). In this way, marginalized social groups can be handi­ capped to the extent that the goal of reaching a consensus decision can exert significant pressures on them to fall in line (Asenbaum 2016). Deveaux claims — quite convincing­ ly — that compromise makes deliberative democracy more receptive to the claims of minority groups(2018, p. 161). But in citizens’ assemblies the point should not be to reach a compromise in the sense of a“deal,” in which the partici­ pants tie up neat negotiation packages containing their re­ spective interests. Instead, a compromise should be the out­ come of a struggle to hammer out a common position from among all of the competing perspectives. It should be rec­ ognized as the best possible solution from an objective point of view, even if it does not completely incorporate everyone’s individual convictions. The crucial point, howev­ er, is this: once a compromise solution has been found, sol­ id reasoning to justify it must be offered to the public. Because under certain circumstances the unanimity require­ ment can generate pressures to conform, nothing really fun­ damental speaks against relying on majority voting. The lat­ ter also mostly conforms to the usual practice in previous cit­ izens’ assemblies. A few authors advocate a qualified form of majority voting to avoid a manipulated consensus or a stale­ mate(Cohen 1989). In that case, 80 percent of the partici­ pants would have to concur with the recommendations (Chwalisz/ Č esnulaityt ė 2020, p. 102). Also, a minority report should be submitted, in which dissenting opinions are aired and the reasons for them made clear. On the one hand, ma­ jority voting poses the danger that“whoever predominates gets to dominate”(Mendelberg et al., p. 23). On the other, it can relieve minorities of the pressure to conform. Whichever decision rule one might choose to implement, in order for cit­ izens’ assemblies to fulfill their mandate, the participants must take their cue from the duties of reciprocal justification, open-endedness, and respect for other positions. All of these obligations should then culminate in analogous decisions. To encourage this action orientation, a clear articulation of these communicative expectations as well as a correspond­ ingly sensitive facilitation of the process will be indispensable. 4.5 AND LEVELING OF STRUCTURAL INEQUALITIES Disadvantaged groups in particular must be included in po­ litical decision-making. But for various reasons this is often not the case. Different starting points as well as structural inequalities of participation play an important role in creat­ ing stumbling-blocks to their inclusion(Beauvais 2018; Schäfer 2010). For people to have equal opportunities to participate in the processes of will-formation and deci­ sion-making, they need sufficient material and cultural re­ sources(Knight/Johnson 1997). And that is exactly why pro­ cedures must be devised for citizens’ assemblies that com­ pensate for the unequal distribution of formal education or enable facilitators to take it fully into account. What is more, we must take seriously the existence of eval­ uative hierarchies, that may, for example, foster a“gender gap” in the deliberative processes(Mendelberg et al. 2014). We have already noted that participants with less formal ed­ ucation may stand somewhat aloof in deliberative arenas. The same thing may happen with women, as relevant stud­ ies have demonstrated(Mansbridge 1983; Gerber 2015). There are also indications that dominant argumentative be­ havior by men can undermine the deliberative capabilities of women(Afsahi 2020). Women seem more inclined to change their views after they have heard the counterargu­ ments of male participants(Beauvais 2019). In other words, a high degree of descriptive(social-structural) representa­ tion does not automatically translate into high substantive representation of the relevant perspectives and needs(Men­ delberg et al. 2014). For these reasons, it is imperative to identify procedural as­ pects of citizens’ assemblies that might aid in reducing tra­ ditional social power disparities. One of those involves the ways in which facilitation or moderation is done. Facilitators (or moderators) must enable an egalitarian discourse that is tailored specifically to the members of disadvantaged groups(Beauvais 2018, p. 151). They have to ensure that in­ dividual participants do not dominate and must guarantee an inclusive debate, taking care that all participants feel comfortable in the discussion.“Facilitator is 90 percent of the quality of your project,” commented Iain Walker in an interview with the authors. Because they are so vital to the success of the project, the standard expectation is that facil­ itators will have undergone professional training. The discursive style of disadvantaged groups often departs from that of the culturally dominant forms of discourse. For that reason, too, such groups often find it difficult to get a hearing in traditional debates(Mendelberg et al. 2014; Polletta/Lee 2006). Consequently, reasoned forms of com­ munication in the traditional sense — such as formal argu­ mentation — should not be the only ones invited and wel­ comed to the deliberation. This narrow or“orthodox” un­ derstanding of deliberation correlates neatly with the habi­ tus — the disposition and social environment — of people with a specific social and educational background, especial­ ly with that of males from the academically trained middle class(Bächtiger et al. 2010; Young 1996). In response, if we want to provide members of other or marginalized groups with equal opportunities to take part, it will prove helpful to encourage a broader, more strongly“ecumenical” under­ standing of deliberation. The latter encompasses a range of communicative forms commonly used in diverse social stra­ ta and milieus(Deveaux 2018; Dryzek et al. 2019). That would go far toward the fuller inclusion of people who are at a disadvantage in conventional forms of democratic par­ ticipation and representation. Thus, deliberation must take into account not only the kind of rationality exhibited in rea­ soned justifications, but also people’s emotions, hopes, and fears. These may find expression in a variety of ways:“story-­ 23

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES telling”(based on personal experiences), input about one’s own perspective and experiences, the choice of words unique to an individual, explicit mutual recognition, and the rhetoric of humor(Gormley 2019). Quite apart from the choice of a communicative style, it is nonetheless crucial that it should be accompanied by the attempt to make one’s own way of seeing understandable and accessible to others, while remaining open to justified objections from the other side. Here, too, the facilitators or moderators become espe­ cially important figures. They must ensure that the setting is integrative and not disrupted or riven by polarization. 4.6 MEDIA COVERAGE Ideally, debates in a citizens’ assembly would be covered by the media, thereby making them accessible to more people than just the ones directly involved. Organizers also see this is a matter of vital significance for the success of the citi­ zens’ assembly(interviews with Claudine Nierth and Iain Walker). According to them, all citizens should be in a posi­ tion to find information about the process easily, whether via online streaming of the plenary sessions or via reporting in the media. Toward that end, personnel should be includ­ ed in the organization who will be responsible for the pub­ lic image of the process of deliberation and who can con­ tribute to the development of a successful communications strategy(Chwalisz 2020, p. 140). It is crucial for the legiti­ mation and acknowledgement of the outcomes of a citi­ zens’ assembly that the public should not have to wait until its decisions are announced to learn in greater detail about its composition and activities. All this should be made pub­ lic before the proceedings actually begin(cf. interview with Iain Walker). Still, it would be unrealistic to assume that tra­ ditional media, from TV to print, will cover every assembly exhaustively. In the best case, they will do their reporting at the beginning and then later when the decisions are an­ nounced. Therefore, online media will need to be empha­ sized much more, since — within certain limits — the organ­ izers themselves can make use of them. They can do a range of things, from disseminating information on their websites to communicating via social media. Because they reach wider audiences, the latter constitute an especially important pillar of media work on citizens’ assemblies. There is an enormous potential inherent in the creative de­ velopment and application of digital tools and platforms. As the example of Taiwan has shown, they can promote broad discursive participation in concrete processes of will-formation and decision-making(Hierlemann/Roch 2020). The issue, significance, and transparency of the pro­ ceedings need to be so interesting for appreciable seg­ ments of the population that they have a powerful motiva­ tion to use the opportunities to acquire information and participate digitally. By contrast, if the broad or at least the interested public is not familiar with the whole process or does not even ap­ prove of it, the citizens’ assembly could be politically margin­ alized. To cite an example, intense media interest compelled Irish politicians to submit the citizens’ assembly’s recommen­ dations to a referendum. As expected, voters who already knew about the assembly’s work supported its recommen­ dations more strongly(Elkin et al. 2016; The Economist 2020). In France, too, public agreement with the recommen­ dations of the French citizens’ assembly on climate(Odoxa 2020) and its familiarity to French citizens(Réseau Action Cli­ mat 2020) put pressure on political actors to implement the proposals(Mellier/Wilson 2020; cf. also interview with An­ toine Vergne). The citizens’ assemblies in Ireland, France, and British Columbia were media fests that may light the way for future media strategies. Nevertheless, prudent“democratic realism” cautions us that such high visibility cannot necessar­ ily be reproduced just anywhere and for any topic. Jane Mansbridge(2018) argues that robust legitimacy also emerges from the public presentation of mini-publics. The media can help to promote broader public deliberation and legitimation. But Mansbridge, a scholar of participation, fails to provide a recipe for communicative success showing how media attention can be garnered for different questions in distinct contexts and at various territorial levels. But at bot­ tom, processes of deliberation and decision-making in citi­ zens’ assemblies can lead to greater acceptance of — even controversial — decisions, since citizens are more likely to trust a decision made — at least in part — by their fellow cit­ izens than one made exclusively by professional politicians, perhaps behind closed doors(Warren/Gastil 2015; interview with Iain Walker). Hence, citizens’ assemblies can act as trusted information proxies for other citizens(Mackenzie/ Warren 2012), assuming that their work is disseminated ef­ fectively. We still do not know empirically exactly what con­ ditions must prevail to produce such an advance in legitima­ cy. Yet apparently this is the case when politically difficult decisions are not made in the“either-or” mode. Instead, what must happen is that, in the aftermath of citizens’ as­ sembly consideration, institutions legitimized by elections such as parliaments and governments will debate the deci­ sions made by the participating citizens and justify their re­ sponse to those decisions to them as well as the general public. It is precisely the complementarity of the two forms of representation that promises a boost in the production of democratic legitimacy. 4.7 CONTROL, AND MONITORING The more influential citizens’ assemblies appear to be, the more enticing are the incentives for political and social ac­ tors to try to manipulate the process from the outside. To prevent special interests and power politics from unduly swaying such assemblies and to make sure they function in a way consistent with their goals, an external, independent body should form a committee that organizes the citizens’ assembly, supervises the process and implements the stated goals. An empirical survey of the OECD shows that, to date, the following non-partisan organizations assumed the role of neutral leadership( Č esnulaityt ė 2020a, p. 78f.): profes­ sional partners from the private sector like Missions Pub­ liques in France, IFOK in Germany, or MASS LBP in Canada, 24

The design of citizens’ assemblies that have specialized in civic participation and deliberation (37 percent); NGOs such as the Australian newDemocracy Foundation or the American Healthy Democracy(29 per­ cent); government organizations like the Danish Board of Technology Foundation or the Office for Future Issues in Vorarlberg, an Austrian federal state, which are non-parti­ san yet still funded by the government(16 percent); partner­ ships between several organizations, for example between a public body and an NGO, an organization from the private sector, or a university institute(10 percent); finally, universi­ ty institutes themselves(8 percent). The organizational committees steer and coordinate the entire process from logistics to the selection of participants, the recruitment and training of facilitators, and the prepa­ ration of the report on final results. To make sure that the requisite competence and neutrality in the selection of ex­ perts and informational materials are preserved, even be­ fore the beginning of the citizens’ assembly’s deliberations, an independent consultative group of scholars and scien­ tists usually is convened( Č esnulaityt ė 2020b). The task of such an advisory board mainly involves offering scientific oversight of the work of the citizens’ assembly and support­ ing it in the selection of the proper experts. However, as noted already, the participants themselves are supposed to attain a certain degree of control over the choice of experts as the process moves ahead. They do so partly by evaluat­ ing critically the expertise on offer and, if need be, calling for additional experts. This procedure increasingly is being taken into account in actual practice as well. Farrell and Suiter(2019) recommend that the body of participants also should form a leadership committee that carries on ex­ changes with both the organizational committee and the advisory board, and represents the concerns of the partici­ pants. This step would preserve the participants’ capacity to act and their authorship at the level of leadership and or­ ganization as well. Different elements of the assembly can parcel out the tasks of control and monitoring among themselves. Although the organizational committee will take care of those matters to some extent, the task also may fall to independent experts, for example to the so-called guarantors in the case of the French climate citizens’ assembly. The latter were named by a variety of organizations(Mellier/Wilson 2020). They were supposed to make sure that the process conformed to nor­ mative rules. Yet because there is no guarantee that inde­ pendent experts really will be neutral and unbiased, Antoine Vergne, in an interview with the authors, suggested that monitoring should be done by a steering group composed of the relevant stakeholders who will have distinctive view­ points on the process. In a few cases the organizers them­ selves took over the monitoring role(interview with Iain Walker). Moreover, another even stronger neutrality check can be guaranteed through the engagement of participat­ ing citizens as exemplified by the Citizen Oversight Group of the planned Citizens’ Convention on UK Democracy(CCUKD 2020). The distribution of monitoring functions to different offices and committees allows the controlling bodies to check up on one another. Beyond what has been stated so far, a scientific evaluation should be conducted in the form of an anonymous partici­ pant questionnaire and, if need be, an expert analysis to gain new knowledge about the process that might be useful for the continued evolution of this form of participation. It is also of enormous importance to keep tabs on the subse­ quent fate of the citizens’ assembly’s recommendations. The one thing that must be prevented is for the citizens’ assem­ bly to serve as a pseudo-participatory placebo and be instru­ mentalized as a toothless legitimation for“business as usu­ al.” For that very reason the permanently institutionalized Citizens’ Dialogue of East Belgium provides for the monitor­ ing both of parliamentary debates and of the steps taken to consider and/or implement agreed-upon measures:“One year at the latest after the final session of the parliamentary committee, the parliamentarians, the minister, and the members of the citizens’ assembly will meet again to discuss where things stand with the implementation of the recom­ mendations”(Citizens’ Dialogue of East Belgium 2021, p. 6). Should it be considered necessary or expedient, additional assemblies can be convened to monitor ongoing implemen­ tation. In other cases, participants already have made use of this option, as for example in the American state of Minne­ sota, where the participants were reconvened to check on how the executive was interpreting their recommendations. And in Australia, workshops were organized to discuss the time horizon for the implementation of a citizens’ assembly’s recommendations(Chwalisz/ Č esnulaityt ė 2020. P. 106f.). 4.8 YSTEMIC EFFICACY AND COMPATIBILITY WITH REPRESENTATIVE DEMOCRACY Democracy is under pressure. That holds true especially for its representative procedures and institutions. Citizens’ trust in parties, parliaments, and governments is languishing at very low levels in the EU member states. The critique of elections as the basic procedure of representative democra­ cy and the accompanying plea for forms of participation based on selection by lot has gained prominence(Buch­ stein/Hein 2009; Manin 2007; van Reybrouck 2016; Gastil/ Wright 2018). Nevertheless, because of the advantages they have from the point of view of democratic theory as well as for political-cultural reasons, universal, free, equal, and fair(as possible) elections and the electoral representa­ tion resulting from them will continue to be crucial, indis­ pensable means of legitimizing political power in democra­ cies(Kneip/Merkel 2017; Warren 2017). So, when we reflect on the democratizing effect of citizens’ assemblies, we always seem to be thinking about some combination of procedures and arenas, not about mutually exclusive alternatives. Thus, the central question is: how can the institutions of electoral representation and the aleato­ ry-participatory instruments complement one another in such a way that democracy as a whole will be strengthened by them? For instance, if in the eyes of the represented the norm-giving institution of parliament is perceptibly losing its power to represent them, whether substantively or descrip­ 25

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES tively, then additional direct-democratic options for partici­ pation may bolster the legitimation of the entire democratic system. That does not mean that citizens’ assemblies can be fitted seamlessly and without tension into existing institu­ tions and political practices. Because of their character as quasi-publics(due to the principles of selection by lot and moderated deliberation), they introduce a new kind of logic into the representative systems, saturated as these are by the logic of competition and influence. Their unique proce­ dures to some extent turn them into competitors of elected representative institutions, but also of parties and interest groups with historically-rooted claims to representation(cf. Dean et al. 2020). Still, the latter should resist the tempta­ tion to instrumentalize citizens’ assemblies as strategic pawns in the political game. Instead, they should attend to those procedures in a constructive but critical spirit, while borrowing from them suggestions for new policy content. Thus, the tension between the two different kinds of logic will linger and should not be dissipated. Without this coop­ erative tension citizens’ assemblies cannot give the neces­ sary push for further democratizing the political system as a whole. If they embed themselves too comfortably in estab­ lished routines and practices, they run the risk of degenerat­ ing into placebo-like instruments for securing legitimacy. In fact, one might observe that we are on the threshold of a trade-off in the present historical situation. On the one hand, citizens’ assemblies can aspire to have an immediate effect by influencing policies; on the other, they can“keep their distance from power” in the sense that they maintain the greatest possible independence from existing electoral majorities. Amid this tension several options emerge for linkages with existing institutions and procedures. Fundamentally, citizens’ assemblies should be systemically effective with their recommendations. The latter should not simply die away without an echo; rather, their substance should be taken up by other institutions and actors within the democratic system. But of course, such efficacy does not come out of thin air. It seems to us that there are certain forms of cooperation that might be especially promising in the sense of their likely impact on the concrete elaboration of policy. This scenario would occur when, let us say, an initia­ tive with either majority or cross-party support came from parliament, the city council, or a government, providing a concrete topic for deliberation in a citizens’ assembly. Gener­ ally speaking, this would guarantee the specific interest of the“sponsor” and increase the likelihood that the assem­ bly’s final recommendations would be taken seriously and then, in whatever form, get translated into a binding decision bearing the imprimatur of the state. In order to avoid leaving this sequence of events solely to the democratic goodwill of the state’s institutions, from the very outset there should be an arrangement between the citizens’ assemblies or their or­ ganizers and the state sponsors that stipulates what their mutual obligations will be. Such agreements should be nego­ tiated and nailed down one by one for every particular citi­ zens’ assembly. Alternatively, in the wake of a few experienc­ es of cooperation the mutual obligations could be prescribed in general laws, the language of which has yet to be formu­ lated. Such a law could specify the duty of the state sponsor to take up the assembly’s recommendations and justify its re­ sponses, or it could go further and insist that the state hold a referendum on the recommendations. But those suggestions do not exhaust the constructive pos­ sibilities inherent in the use of citizens’ assemblies. A“bot­ tom-up variant” of the assemblies certainly ought to supple­ ment the“top-down variant” described above. That would be the case if and when citizens gathered a certain number of signatures on a petition to the above-mentioned institu­ tions proposing a concrete topic for a citizens’ assembly and requiring the latter to take a position on the proposals. As the exact topic is thrashed out, modifications still should be possible to ensure that the representative institutions are able to cooperate. In case the legislative or executive com­ pletely rejects the proposals, they would have to justify their decision in a formal, transparent, and legally valid manner. This category also should include the possibility that a par­ liamentary minority(Setälä 2020) acting as a“deliberative opposition”(Schäfer 2020) could initiate a citizens’ assem­ bly. Also, in the context of these options, it seems reasona­ ble to impose on the parliament or government the duty to respond formally to the recommendations of the citizens’ assembly that have been addressed to them, either by ac­ cepting them wholesale or having to defend publicly their (partial) rejection. Furthermore, if a given constitution al­ lowed for popular initiatives or plebiscites, it would make sense to hold a mandatory citizens’ assembly in advance. The assembly could either formulate recommendations for the voting decision or even be granted the right to reword the question to be voted on. Quite apart from the issue of how closely linked to current political majorities an initiative may be, it appears funda­ mentally desirable that citizens’ assemblies should not be es­ tablished alongside representative institutions but instead brought into close contact with them. This holds true for the connection between citizens’ assemblies and parliaments and executives from local government on up to the federal states and from there to the federal government(the central state) itself. Such a“linked-in” procedure would make it possible to produce a dual legitimation for the democratiza­ tion of democracy. For one thing the citizens’ assemblies would acquire a higher public profile and stir up greater en­ thusiasm for participation amongst the populace. For that to happen it must be clear that the recommendations have meaning in the real world of politics and have not just been exploited by the state authorities as a participatory placebo. Under those conditions elected decision-making elites and representative institutions would take the citizens’ assem­ blies seriously. For another, state institutions would bolster their stock of legitimacy among the citizenry since they could make it clear that they had an interest in the opinions and cooperation of average citizens and that they would not act as a political class isolated from the rest of society, rep­ resenting only their own interest and perspectives. The citi­ zens’ ties to the commonwealth would deepen since they could now directly and responsibly help shape the policies under which they have to live. Both sides would be the win­ ners, as would democracy as a whole. 26

To generate this democratic bonus in a relevant dimension, citizens’ assemblies would have to be established at every level of the political system, from municipalities and coun­ ties or districts all the way up to the nation-state(and pos­ sibly beyond). Unique gems such as the nationally signifi­ cant Citizens’ Assemblies in Ireland and Canada function as highly visible beacons on the international scene. But citi­ zens’ assemblies will not reveal their true significance for democracy as such until they can be“rolled out” far and wide. The design of citizens’ assemblies 27

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES 5 CONCLUSION AND OUTLOOK FOR THE FUTURE What is the democratization potential of citizens’ assem­ blies? What challenges lurk along the road to its implemen­ tation? Can citizens’ assemblies offer a constructive re­ sponse to the problems and crisis phenomena that current­ ly plague representative democracy, such as dwindling trust, selective non-participation, and social and political polariza­ tion? The answer is a conditional“yes.” On the one hand, citizens’ assemblies can help bridge the perceived gap between average citizens and political deci­ sion-makers. By combining stratified random selection with deliberative procedures, it is possible to include many more people from highly diverse social, political, and cul­ tural backgrounds. As a specific form of deliberative mi­ ni-publics they import a new logic into a representative system marked by electoral competition. In that way many of the disadvantages of elections can be defused to a con­ siderable extent. The list of such disadvantages is lengthy: the focus on elites; lagging voter turnout specific to the lower(educational) classes; the eclipse of substantive poli­ cy matters by political marketing; and the preoccupation with the present at the expense of future issues which tends to be induced by the electoral cycle. Moreover, as empirical and theoretical studies have shown, citizens’ as­ semblies can achieve a variety of concrete purposes while successfully managing a highly diverse set of issues — often the most difficult and controversial ones. Those accom­ plishments are not trivial in an age of increasing social and political polarization. By contributing to the acknowledge­ ment of the legitimacy of different preferences and opin­ ions, they counteract the spread of friend-enemy relation­ ships and strengthen social cohesion. On the other hand, citizens’ assemblies cannot exploit this potential fully unless they fulfill three key criteria: descriptive inclusion, high deliberative quality, and systemic efficacy. First, the drawing of lots and invitation procedures must be set up in such a way that, in fact, people from all affected social strata take part. Second, the internal communications process must be organized and facilitated so that all partici­ pants, regardless of their social background and habitus, can get involved as equals and be heard. Furthermore, the outcome of the process itself must be open-ended. This structure also implies that the citizens’ assembly or a third, independent body in addition to the initiator will have suffi­ cient control over the assembly’s agenda. Third, citizens’ as­ semblies must be embedded in the democratic political sys­ tem as an autonomous and influential factor. None of these things will happen as a matter of course, be­ cause democratic innovations like citizens’ assemblies, not to mention the existing institutions of representative de­ mocracy, must face the same set of daunting social condi­ tions from the very beginning, including especially growing social inequality, which is reflected increasingly in political inequality as well. In addition, political actors competing for attention and votes have incentives to instrumentalize citi­ zens’ assemblies as legitimizers of existing projects, which in turn would jeopardize their autonomy. To maintain their democratizing effect even against those odds, citizens’ as­ semblies will have to follow a careful design, one that lays out a set of transparent objectives to guide their proceed­ ings. The targets thus chosen also should respect the open­ ness of the decision-making process and be based on a clear justification. If they follow those guidelines, citizens’ assem­ blies will maximize their chances to live up to the normative expectations placed on them and effectively to inject the perspectives of the average citizen into the political process. When citizens and the broader public perceive that citizens’ assemblies keep their promises, trust in political participation in the democratic process may grow. In that case short-, mid­ dle-, and long-term prospects for their evolution start to emerge. Currently, and focusing now on the short run, we can identify a phase of relative open-endedness and experi­ mentation with the format that has brought forth a variety of experiences in recent years. In the middle term — and building on these experiences and learning processes — we should transition to a new phase in which citizens’ assem­ blies are institutionalized gradually. The effort to give them a consistent, regular status can happen at different levels with­ in the state and in similar or perhaps more sophisticated ways. In the long run one may hope that citizens’ assemblies become part of a transformation of the political culture. If citizens’ assemblies were to become a self-evident compo­ nent of the process of democratic will-formation alongside other forms of participation, and were deployed on a contin­ uing and not merely ad hoc basis, then most citizens could count on being selected by lot to take part in a citizens’ as­ sembly at least once in their lives at some political level. Pro­ vided that all those conditions were met, this form of partic­ ipation eventually could develop a formidable democra­ 28

cy-enhancing power of integration. If such a cultural shift were to succeed, one might also imagine that the by now well-tested modes of selection and procedure would bestow so much legitimacy that citizens’ assemblies — in addition to other elected bodies — would be entrusted with far-reaching decision-making responsibilities. The latter might extend to the right to take up issues on their own authority, exercise veto options, claim an initiative right, and even share in leg­ islative decisions as a second or third chamber. Whether or not such paths into the future are taken depends on decisions being made right now. Along the way, the clas­ sical institutions and actors of representative democracy have the opportunity to use democratic innovations like citi­ zens’ assemblies to reflect upon, cooperatively strengthen, and partially redefine their own roles. So, it is not a question of crowding out or disempowering classical representative institutions, but of making them part of a more effective democratic ensemble. Citizens’ assemblies are comple­ ments, not substitutes. Given the challenges of the present, representative democracy will be able to secure its future permanently only if the main political decision-makers con­ stantly try to innovate, aiming to approach more closely the democratic promise of participation(in principle) by all. Citi­ zens’ assemblies can contribute to that effort. Conclusion and outlook for the future 29

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES HOW TO MAKE A CITIZENS’ ASSEMBLY SUCCEED Steps and phases Actors and arenas Activities and processes Criteria and quality control Initiation and agenda-setting Civil society/signature petitions; parliamentary opposition; govern­ ment or parliamentary majority; parties Soliciting public support; proposal of topics; occasional negotiations between societal and state actors Relevant political issues; clear and binding objectives; in some cases, obligation of commitment of state actors to consider the outcomes; appropriate budgeting and time frames Selection and recruitment NGO, professional participation or market research institute, media Stratified random selection, solic­ ited participation; public informa­ tion campaigns; media coverage Focus on thematically relevant characteristics, diversity, marginal­ ized perspectives; public attention Learning Citizens’ assembly: plenary session or, if occasion arises, hybrid for­ mats with digital presentations, in­ put-, storage- and retrieval options Constitutive sessions; presentations and question-and-answer sessions; revision of the agenda by the citi­ zens’ assembly; public streaming of the expert presentations Neutrality or diversity of informa­ tion and sources; information trans­ fer in generally understandable lan­ guage; critical thinking; control options for participants and inde­ pendent monitoring Consulting Members of the citizens’ assembly, supporting institutes; stakehold­ ers, broader public; analog and dig­ ital forums Public hearings as needed, regional conferences, town-hall meetings Arranging an exchange with the broader public; plurality and balance, inclusion of marginalized perspectives Deliberating Citizens’ assembly, plenary meet­ ings and small groups; facilitators Moderated plenary and small group discussions; reflection, revi­ sion, and elaboration of positions; monitoring of the facilitators Clear and sufficient time horizon; respectful, uncoerced and inclu­ sive deliberation; professional train­ ing of the facilitators;“ecumenical” understanding of deliberation Deciding Citizens’ assembly, small groups and plenary meeting; media Formulation of and voting on rec­ ommendations; publicity work and media coverage Clear rules governing decisions; uncoerced consensus or fair and appropriate compromise; wellfounded minority votes when ma­ jority has decided; duty of justifi­ cation Uptake/ consideration of the results Political decision-makers, members of the assembly; parties; media Submission of the results; referen­ dum or parliamentary proceedings (acceptance or reasoned rejection); monitoring Feedback, justification, reciprocal exchange between citizens’ assem­ bly and decision-makers 30

REFERENCES REFERENCES Afsahi, Afsoun(2020): Gender Difference in Willingness and Capacity for Deliberation. Social Politics: International Studies in Gender, State& Society 2020: jxaa003. doi:10.1093/sp/jxaa003. Asenbaum, Hans(2016): Facilitating Inclusion: Austrian Wisdom Coun­ cils as Democratic Innovation between Consensus and Diversity. Journal of Public Deliberation 12(2): Article 7. doi:10.16997/jdd.259. Bächtiger, André, Maija Setälä, and Kimmo Grönlund(2014): To­ wards a New Era of Deliberative Mini-Publics. In Deliberative mini-publics. Involving citizens in the democratic process, Edited by Kimmo Grönlund, André Bächtiger und Maija Setälä: pp. 225–246. Colchester: ECPR Press. Bächtiger, André, Simon Niemeyer, Michael Neblo, Marco R. Steenbergen, and Jürg Steiner(2010): Disentangling diversity in de­ liberative democracy. Competing theories, their blind spots and comple­ mentarities. Journal of Political Philosophy 18(1): pp. 32–63. doi:10.1111/ j.1467-9760.2009.00342.x. Barber, Benjamin R.(1994): Starke Demokratie. Über die Teilhabe am Politischen. Hamburg: Rotbuch-Verlag. Beauvais, Edana(2018): Deliberation and Equality. In The Oxford Hand­ book of Deliberative Democracy, Edited by Andre Bächtiger, John S. Dry­ zek, Jane Mansbridge and Mark Warren: pp. 144–155. Oxford: Oxford University Press. Beauvais, Edana(2019): Discursive inequity and the internal exclusion of women speakers. Political Research Quarterly 74(1): pp. 103–116. doi:10.1177/1065912919870605. Böker, Marit(2017): Justification, critique and deliberative legitimacy: The limits of mini-publics. Contemporary Political Theory 16(1): pp. 19–40. doi:10.1057/cpt.2016.11. Buchstein, Hubertus(2009): Demokratie und Lotterie. Das Los als poli­ tisches Entscheidungsinstrument von der Antike bis zur EU. Frankfurt/ Main: Campus. Buchstein, Hubertus, and Michael Hein(2009): Randomizing Europe: the lottery as a decision-making procedure for policy creation in the EU. Critical Policy Studies 3(1): pp. 29–57. doi:10.1080/19460170903158081 Bürgerdialog in Ostbelgien(2021): Die wichtigsten Antworten zum Ablauf des Bürgerdialogs. https://www.buergerdialog.be/fileadmin/user_ upload/Buergerdialog-Ablaufbeschreibung_V20201106.pdf. Accessed 17.08.2021. Caluwaerts, Didier, and Min Reuchamps(2016): Generating Demo­ cratic Legitimacy through Deliberative Innovations: The Role of Embed­ dedness and Disruptiveness. Representation 52(1): pp. 13–27. doi:10.108 0/00344893.2016.1244111. CCUKD(2020): A Citizens’ Convention On UK Democracy. https://static1. squarespace.com/static/5d31c7fcfab26d00013cca0f/t/5f7b82c409e2b­ 5141f2e3a73/1601929948359/CCUKD+PROPOSAL+OCT+2020_FA_DIG­ ITAL.pdf. Accessed: 17.08.2021. Centre for Public Impact(2019): Estonia Citizens’ Assembly, Restor­ ing Political Legitimacy. https://www.centreforpublicimpact.org/casestudy/estonia-citizens-assembly-restoring-political-legitimacy. Accessed: 17.08.2021. Č esnulaityt ė , Ieva(2020a): Key trends. In Innovative Citizen Participa­ tion and New Democratic Institutions: Catching the Deliberative Wave, Edited by OECD, p. 65–79. Paris: OECD Publishing. Č esnulaityt ė , Ieva(2020b): Models of representative deliberative pro­ cesses. In Innovative Citizen Participation and New Democratic Institu­ tions: Catching the Deliberative Wave, Edited by OECD, pp. 33–64. Paris: OECD Publishing. Chambers, Simone(2003): Deliberative Democratic Theory. Annual Review of Political Science 6(1): pp. 307–326. doi:10.1146/annurev. polisci.6.121901.085538. Chambers, Simone(2009): Rhetoric and the Public Sphere: Has Delib­ erative Democracy Abandoned Mass Democracy? Political Theory 37(3): pp. 323–350. doi:10.1177/0090591709332336. Chwalisz, Claudia(2020): Reimagining democratic institutions: Why and how to embed public deliberation. In Innovative Citizen Participation and New Democratic Institutions: Catching the Deliberative Wave, Edited by OECD, pp. 121–148. Paris: OECD Publishing. Chwalisz, Claudia, and Ieva Č esnulaityt ė (2020): What is a‘successful’ representative deliberative process for public decision making? Assessing the evidence. In Innovative Citizen Participation and New Democratic Insti­ tutions: Catching the Deliberative Wave, Edited by OECD, pp. 80–114. Paris: OECD Publishing. Cohen, Joshua(1989): Deliberation and Democratic Quality. In The Good Polity: Normative Analysis of the State, Edited by Alan Hamlin and Phillip Petit: pp. 17–34. Oxford/New York: Blackwell. Convention citoyenne de Nantes(2021): Convention citoyenne de Nantes: les citoyens rendent leur avis. https://conventioncitoyenne-nant­ esmetropole.fr/11-mars/. Accessed: 17.08.2021. Courant, Dimitri(2021): The promises and disappointments of the French Citizens’ Convention for Climate. https://www.publicdeliberation. net/the-promises-and-disappointments-of-the-french-citizens-conven­ tion-for-climate/. Accessed: 17.08.2021. Curato, Nicole, John S. Dryzek, Selen A. Ercan, Carolyn M. Hendriks, and Simon Niemeyer(2017): Twelve Key Findings in Deliberative Democracy Research. Daedalus 146(3): pp. 28–38. Curato, Nicole, David M. Farrell, Brigitte Geissel, Kimmo Grönlund, Patricia Mockler, Jean-Benoit Pilet, Alan Renwick, Jonathan Rose, Maija Setälä, and Jane Suiter(2021): Deliberative Mini-Publics. Core Design Features. Bristol: Bristol University Press. Dahl, Robert A.(1989): Democracy and its critics. New Haven: Yale Uni­ versity Press. Dean, Rikki, John Boswell, und Graham Smith(2020): De­ signing Democratic Innovations as Deliberative Systems: The Am­ bitious Case of NHS Citizen. Political Studies 68(3): pp. 689–709. doi:10.1177/0032321719866002. Deveaux, Monique(2018): Deliberative Democracy and Multicultural­ ism. In The Oxford Handbook of Deliberative Democracy, Edited by André Bächtiger, John S. Dryzek, Jane Mansbridge and Mark Warren, pp. 156–170. Oxford: Oxford University Press. Druckmann, James, Erik Peterson, and Rune Slothuus(2013): How Elite Partisan Polarization Affects Public Opinion Formation. American Political Science Review 107(1): pp. 57–79 doi:10.1017/ S0003055412000500. Dryzek, John S.(2000): Deliberative Democracy and Beyond. Liberals, Critics, Contestations. Oxford/New York: Oxford University Press. Dryzek, John S., André Bächtiger, Simone Chambers, Joshua Cohen, James N. Druckman, Andrea Felicetti, James S. Fishkin, David M. Farrell, Archon Fung, Amy Gutmann, Hélène Landemore, Jane Mansbridge, Sofie Marien, Michael A. Neblo, Simon Niemeyer, Maija Setälä, Rune Slothuus, Jane Suiter, Dennis Thompson, and Mark E. Warren(2019): The crisis of democracy and the sci­ ence of deliberation. Science 363(6432): pp. 1144–1146. doi:10.1126/ science.aaw2694. Elkin, Johan A., David M. Farrell, Theresa Reidy, and Jane Suiter (2016): Understanding the 2015 marriage referendum in Ireland: context, campaign and conservative Ireland. Irish Political Studies 32(3): pp. 361–381. doi:10.1080/07907184.2016.1197209. Elstub, Stephen(2014): Mini-publics: Issues and Cases. In Delibera­ tive democracy. Issues and cases, Edited by Stephen Elstub and Peter McLaverty, pp. 166–188. Edinburgh: Edinburgh University Press. 31

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES Elstub, Stephen, and Oliver Escobar(2019): Defining and typologising democratic innovations. In Handbook of Democratic Innovation and Gov­ ernance, Edited by Stephen Elstub and Oliver Escobar, p. 11–32. Chelten­ ham/Northampton: Edward Elgar Publishing. Escobar, Oliver, and Stephen Elstub(2017): Forms of Mini Publics: An Introduction to Deliberative Innovations in Democratic Practice. new­ DEMOCRACY 2017(4): pp. 1–14. https://www.newdemocracy.com. au/2017/05/08/forms-of-mini-publics/. Accessed: 17.08.2021. Eriksen, Erik O.(2009): The Unfinished Democratization of Europe. Ox­ ford: Oxford University Press. Farrell, David M., Nicole Curato, John S. Dryzek, Brigitte Geißel, Kimmo Grönlund, Sofie Marien, Simon Niemeyer, Jean-Benoit Pilet, Alan Renwick, Jonathan Rose, Maija Setälä, and Jane Suiter (2019): Deliberative Mini-Publics: Core Design Features. The Centre for Deliberative Democracy& Global Governance, Institute for Governance and Policy Analysis, University of Canberra. ACT 2601(Working Paper Series No. 2019/5). Farrell, David M., and Jane Suiter(2019): Reimagining Democracy. Lessons in Deliberative Democracy from the Irish Front Line. Ithaca/Lon­ don: Cornell University Press. Farrell, David M., and Peter Stone(2020): Sortition and Mini-Publics. In The Oxford Handbook of Political Representation in Liberal Democra­ cies, Edited by Robert Rohrschneider and Jacques J. A. Thomassen, p. 227–246. Oxford: Oxford University Press. Font, Joan, Graham Smith, Carol Galais, and Pau Alarcon(2018): Cherry-Picking Participation: Explaining the Fate of Proposals from Partici­ patory Processes. European Journal of Political Research 57(3): p. 615–636. Fournier, Patrick, Henk van der Kolk, R. Kenneth Carty, André Blais, and Jonathan Rose(2011): Why Citizen Assemblies and How did they Work? In When citizens decide. Lessons from citizens’ assemblies on electoral reform, Edited by Patrick Fournier, pp. 21–50. Oxford: Oxford University Press. Gastil, John, and Erik Olin Wright(2018): Legislature by Lot: Envisioning Sortition within a Bicameral System. Politics& Society 46(3): pp. 303–330. doi:10.1177/0032329218789886. Gerber, Marlène(2015): Equal Partners in Dialogue? Participation Equal­ ity in a Transnational Deliberative Poll(Europolis). Political Studies 63(1): pp. 110–130. doi:10.1111/1467-9248.12183. Geißel, Brigitte, and Stefan Jung(2019): Mehr Mitsprache wagen. Ein Beteiligungsrat für die Bundesrepublik. http://library.fes.de/pdf-files/ fes/15848.pdf. Accessed: 18.08.2021. Geißel, Brigitte, Rikki Dean, Stefan Jung, and Bruno Wipfler(2019): Bürgerrat Demokratie. Abschlussbericht der wissenschaftlichen Evalua­ tion. Forschungsstelle Demokratische Innovationen, Goethe-Universität Frankfurt am Main. https://www.buergerrat.de/fileadmin/downloads/eval­ uationsbericht.pdf. Accessed: 18.08.2021. Goldberg, Saskia(2021): Just Advisory and Maximally Representative: A Conjoint Experiment on Non-Participants’ Legitimacy Perceptions of Deliberative Forums. Journal of Deliberative Democracy 17(1): pp. 56–75. doi:10.16997/jdd.973. Gormley, Steven(2019): Deliberation, Unjust Exclusion, and the Rhetori­ cal Turn. Contemporary Political Theory 18(2): pp. 202–226. doi:10.1057/ s41296-018-0254-z. Goodin, Robert E., and John S. Dryzek(2006): Deliberative Impacts: The Macro-Political Uptake of Mini-Publics. Politics& Society 34(2): pp. 219–244. doi:0.1177/0032329206288152. Gutmann, Amy, and Dennis Thompson(2018): Reflections on Deliber­ ative Democracy: When Theory Meets Practice. In The Oxford Handbook of Deliberative Democracy, Edited by Andre Bächtiger, John S. Dryzek, Jane Mansbridge and Mark Warren, pp. 900–912. Oxford: Oxford University Press. Habermas, Jürgen(1994): Faktizität und Geltung. Beiträge zur Diskurs­ theorie des Rechts und des demokratischen Rechtsstaats. Frankfurt am Main: Suhrkamp. Habermas, Jürgen(1996): Die Einbeziehung des Anderen. Studien zur politischen Theorie. Frankfurt am Main: Suhrkamp. Hanson, Rebecca(2018): Deepening Distrust: Why Participatory Exper­ iments Are Not Always Good for Democracy. The Sociological Quarterly 59(1): pp. 145–167. doi:10.1080/00380253.2017.1383145. Harris, Clodagh(2019): Mini-publics: design choices and legitimacy. In Handbook of Democratic Innovation and Governance, Edited by Stephen Elstub and Oliver Escobar, pp. 45–59. Cheltenham/Northampton: Edward Elgar Publishing. Hendriks, Carolyn M.(2016): Coupling citizens and elites in deliberative systems. The role of institutional design. European Journal of Political Re­ search 55(1): pp. 43–60. doi:10.1111/1475-6765.12123. Hierlemann, Dominik, and Stefan Roch(2020): Digital Democracy What Europe can learn from Taiwan. Gütersloh: Bertelsmann Stiftung. https://www.bertelsmann-stiftung.de/fileadmin/files/BSt/Publikationen/ GrauePublikationen/Digital_Democracy_-_What_Europe_can_learn_ from_Taiwan.pdf. Accessed: 18.08.2021. Jacquet, Vincent(2017): Explaining non-participation in deliberative mini-publics. European Journal of Political Research 56(3): pp. 640–659. doi:10.1111/1475-6765.12195. James, Rabinder Michael(2008): Descriptive representation in the British Columbia Citizens’ Assembly. In Designing deliberative democ­ racy: The British Columbia Citizens’ Assembly, Edited by Mark Warren and Hilary Pearse, pp. 106–126. Cambridge: Cambridge University Press. doi:10.1017/CBO9780511491177.007. Jonsson, Magnus E.(2015): Democratic Innovations in Deliberative Sys­ tems – The Case of the Estonian Citizens’ Assembly Process. Journal of Public Deliberation 11(1): Article 7. doi:10.16997/jdd.224. Karlsson, Martin, Jonsson Magnus, and Joachim Aström(2015): Did the Estonian Citizens’ assembly help restore political legitimacy?: Analyzing changes in vertical and horizontal trust among participants. Artikel präsentiert auf der ECPR General Conference in Montreal 2015. http://www.diva-portal.org/smash/get/diva2:853063/FULLTEXT01.pdf. Accessed: 18.08.2021. Karpowitz, Christopher F., Chad Raphael, and Allen S. Hammond (2009): Deliberative Democracy and Inequality: Two Cheers for Enclave Deliberation among the Disempowered. Politics& Society 37(4), pp. 576–615. doi:10.1177/0032329209349226. Kneip, Sascha, and Wolfgang Merkel(2017): Garantieren Wahlen demokratische Legitimität? Aus Politik und Zeitgeschichte 67(38–39/2017): pp. 18–24. Kneip, Sascha, and Wolfgang Merkel(2020): Demokratische Legiti­ mität: Ein theoretisches Konzept in empirisch analytischer Absicht. In Le­ gitimitätsprobleme. Zur Lage der Demokratie in Deutschland, Edited by Sascha Kneip, Wolfgang Merkel and Bernhard Weßels, pp. 25–55. Wies­ baden: Springer VS. Knight, Jack, and James Johnson(1997): What Sort of Political Equal­ ity Does Deliberative Democracy Require?. In Deliberative Democracy. Es­ says on Reason and Politics, Edited by James Bohman and Wiliam Rehg, pp. 279–319. Cambridge, MA: MIT Press. Kübler, Lukas, Claus Leggewie, and Patrizia Nanz(2021): Demokra­ tische Innovation durch Bürgerräte. Aus Politik und Zeitgeschichte 71 (26–27/2021): pp. 47–53. Lafont, Cristina(2015): Deliberation, Participation, and Democratic Le­ gitimacy: Should Deliberative Mini-Publics Shape Public Policy?. Journal of Political Philosophy 23(1), pp. 40–63. doi:10.1111/jopp.12031. Lafont, Cristina(2020): Democracy without Shortcuts. A Participatory Conception of Deliberative Democracy. Oxford: Oxford University Press. Lehmann, Pola, Sven Regel, and Sara Schlote(2015): Ungleichheit in der politischen Repräsentation. Ist die Unterschicht schlechter repräsenti­ ert?. In Demokratie und Krise. Zum schwierigen Verhältnis von Theo­ rie und Empirie, Edited by Wolfgang Merkel, pp. 157–180. Wiesbaden: Springer VS. 32

REFERENCES List, Christian, Robert C. Luskin, James S. Fishkin, and Iain McLean (2013): Deliberation, Single-Peakedness, and the Possibility of Meaning­ ful Democracy: Evidence from Deliberative Polls. The Journal of Politics 75 (1): pp. 80–95. doi:10.1017/S0022381612000886. MacKenzie, Michael K., and Mark E. Warren(2012): Two TrustBased Uses of Minipublics in Democratic Systems. In Deliberative Sys­ tems: Deliberative Democracy at the Large Scale, Edited by John Par­ kinson and Jane Mansbridge, pp. 95–124. Cambridge: Cambridge University Press. MacKenzie, Michael K.(2016): Institutional design and sources of short-termism. In Political Institutions for Future Generations, Edited by Iñigo González-Ricoy and Axel Gosseries, pp. 24–46. Oxford: Oxford Uni­ versity Press. Manin, Bernard(1997): The Principles of Representative Government. Cambridge/New York: Cambridge University Press. Manin, Bernard(2007): Kritik der repräsentativen Demokratie. Berlin: Matthes& Seitz. Mansbridge, Jane J.(1983): Beyond Adversary Democracy. Chicago: University of Chicago Press. Mansbridge, Jane(2018): Deliberative Polling Comes of Age. The Good Society 27(1-2): pp. 118–129. doi:10.5325/goodsociety.27.1-2.0118. Mansbridge, Jane, Janette Hartz-Karp, Matthew Amengual, and John Gastil(2006): Norms of Deliberation: An Inductive Study. Journal of Public Deliberation 2(1): Article 7. doi:10.16997/jdd.35. Mansbridge, Jane, James Bohman, Simone Chambers, David Est­ lund, Andreas Føllesdal, Archon Fung, Cristina Lafont, Bernard Manin, and José Luis Martí(2010): The Place of Self-Interest and the Role of Power in Deliberative Democracy. The Journal of Political Philoso­ phy 18(1): pp. 64–100. Mansbridge, Jane, James Bohman, Simone Chambers, Thomas Christiano, Archon Fung, John R. Parkinson, Dennis F. Thompson, and Mark E. Warren(2012): A systemic approach to deliberative de­ mocracy. In Deliberative Systems, Edited by James Parkinson and Jane Mansbridge, pp. 1–26. New York: Cambridge University Press. Meiering, David, Andreas Schäfer(2020):(Ent-)Politisierung- Debat­ ten, Modelle und Befunde. In(Ent-)Politisierung? Die demokratische Ge­ sellschaft im 21. Jahrhundert. Leviathan – Berliner Zeitschrift für Sozial­ wissenschaft(Sonderband 35), Edited by Andreas Schäfer, and David Meiering, pp. 11–36. Baden-Baden: Nomos. Mellier-Wilson, Claire(2020): Convention Citoyenne Pour Le Climat: What Can We Learn From The French Citizens’ Assembly on Climate Change? https://www.involve.org.uk/resources/blog/opinion/conventioncitoyenne-pour-le-climat-what-can-we-learn-french-citizens. Accessed: 18.08.2021. Mellier, Claire, and Rich Wilson(2020): Getting Climate Citizens’ Assemblies Right. https://carnegieeurope.eu/2020/11/05/getting-­climatecitizens-assemblies-right-pub-83133. Accessed: 18.08.2021. Mendelberg, Tali, Christopher Karpowitz, and Baxter Oliphant (2014): Gender Inequality in Deliberation: Unpacking the Black Box of Interaction. Perspectives on Politics 12(1): pp. 18–44. doi:10.1017/ S153759271300369 Mercier, Hugo, and Hélène Landemore(2012): Reasoning Is for Ar­ guing. Understanding the Successes and Failures of Deliberation. Political Psychology 33(2), pp. 243–258. Merkel, Wolfgang(2015): Demokratie und Krise. Zum schwierigen Ver­ hältnis von Theorie und Empirie. Wiesbaden: Springer VS. Merkel, Wolfgang(2016): Krise der Demokratie? Anmerkungen zu ei­ nem schwierigen Begriff. Aus Politik und Zeitgeschichte 66(40–42): pp. 4–11. Merkel, Wolfgang, and Claudia Ritzi(Edited by)(2017): Die Legitim­ ität direkter Demokratie. Wie demokratisch sind Volksabstimmungen? Wiesbaden: Springer VS. Merkel, Wolfgang, and Sascha Kneip(Edited by)(2018): Democracy and Crisis. Challenges in Turbulent Times. Cham: Springer International. Merkel, Wolfgang(2019): Past, Present and Future of Democracy. Pol­ icy Review. European Commission. https://op.europa.eu/en/publica­ tion-detail/-/publication/4bebf83d-60ba-11e9-b6eb-01aa75ed71a1. Accessed: 18.08.2021. Merkel, Wolfgang(2020): Demokratiekrisen. In Handbuch Krisen­ forschung, Edited by Frank Bösch, Nicole Deitelhoff, and Stefan Kroll, pp. 111–113. Wiesbaden: Springer VS. Merkel, Wolfgang(2021): Wissenschaft, Moralisierung und die Demokratie, Aus Politik und Zeitgeschichte(26–27/2021): pp. 4–10. Merkel, Wolfgang, and Anna Lührmann(2021): Resilience of democ­ racies: responses to illiberal and authoritarian challenges. Democratization, 28(5): pp. 869–884. doi:10.1080/13510347.2021.1928081. Nanz, Patrizia, and Claus Leggewie(2018): Die Konsultative: Mehr Demokratie durch Bürgerbeteiligung. Berlin: Wagenbach. Neblo, Michael A., Kevin M. Esterling, Ryan P. Kennedy, David M. J. Lazer, and Anand E. Sokhey(2010): Who wants to delib­ erate- and why?. American Political Science Review 104(3): pp. 566–583. doi:10.1017/S0003055410000298. Neblo, Michael A.(2015): Deliberative Democracy between Theory and Practice. Cambridge: Cambridge University Press. Niemeyer, Simon, John S. Dryzek(2007): The Ends of Delib­ eration. Meta-consensus and Inter-subjective Rationality as Ideal Outcomes. Swiss Political Science Review 13(4), pp. 497–526. doi:10.1002/j.1662-6370.2007.tb00087.x. Odoxa(2020): Les mesures de la Convention citoyenne séduisent les Français… à l’exception notable des 110 km/h. http://www.odoxa.fr/ sondage/mesures-de-convention-citoyenne-seduisent-francais-a-lexcep­ tion-notable-110-km-h/. Accessed: 18.08.2021. Parkinson, John, and Jane Mansbridge(2012): Deliberative systems: Deliberative democracy at the large scale. Cambridge: Cambridge Univer­ sity Press. Polletta, Francesca, and John Lee(2006): Is Telling Stories Good for Democracy? Rhetoric in Public Deliberation after 9/11. American Socio­ logical Review 71(5): pp. 699–723. Rahvakogu(2014): About the Estonian people’s assembly in 2013. https://rahvakogu.ee/peoples-assembly-in-2013/. Accessed: 18.08.2021. Reseau Action Climat(2020): Sondage: Des Gaulois Pas Si Réfractaires A L’Action Climatique. https://reseauactionclimat.org/sondage-des-gau­ lois-pas-si-refractaires-a-laction-climatique/. Accessed: 18.08.2021. Roberts, Jennifer J., Ruth Lightbody, Ragne Low, and Stephan Elstub(2020): Experts and evidence in deliberation: scrutinising the role of witnesses and evidence in mini-publics, a case study. Policy Sciences 53(1): pp. 3–32. doi:10.1007/s11077-019-09367-x. Ryan, Matthew, Graham Smith(2014): Defining Mini-Publics. In Delib­ erative mini-publics. Involving citizens in the democratic process, Edited by Kimmo Grönlund, André Bächtiger, and Maija Setälä, pp. 9–26. Col­ chester: ECPR Press. Schäfer, Andreas(2017a): Zwischen Repräsentation und Diskurs. Zur Rolle von Deliberation im parlamentarischen Entscheidungsprozess. Wies­ baden: Springer VS. Schäfer, Andreas(2017): Deliberation in representative institutions. An analytical framework for a systemic approach. Australian Journal of Politi­ cal Science 52(3), pp. 419–435. doi:10.1080/10361146.2017.1330397. Schäfer, Andreas(2020): Deliberative Opposition? Der parlamentarische Beratungsprozess im Spannungsfeld zwischen Deliberation und Dualis­ mus. In Kritik, Kontrolle, Alternative. Was leistet die parlamentarische Opposition?, Edited by Stephan Bröchler, Manuela Glaab, and Helmar Schöne: pp. 73–91. Wiesbaden: Springer VS. 33

FES DEMOCRACY OF THE FUTURE – CITIZENS’ ASSEMBLIES Schäfer, Andreas, and Wolfgang Merkel(2020): Emanzipation oder Reaktion: Wie konservativ ist die deliberative Demokratie? Politische Vi­ erteljahresschrift 61: pp. 449–472. doi:10.1007/s11615-020-00232-8. Schäfer, Andreas, and Wolfgang Merkel(2021): The Temporal Con­ stitution of Democracies. In The Oxford Handbook of Time and Poli­ tics, Edited by Klaus H. Goetz. Oxford: Oxford University Press.(forth­ coming). Schäfer, Armin(2010): Die Folgen sozialer Ungleichheit für die Demokratie in Westeuropa. Zeitschrift für vergleichende Politikwissen­ schaft 4(1): pp. 131–156. doi:10.1007/s12286-010-0075-9. Setälä, Maija(2017): Connecting Deliberative Mini-Publics to Repre­ sentative Decision Making. European Journal of Political Research 56(4): p. 846–863. doi:10.1111/1475-6765.12207. Setälä, Maija, and Graham Smith(2018): Mini-Publics and Deliberative Democracy. In The Oxford Handbook of Deliberative Democracy, Edited by Andre Bächtiger, John S. Dryzek, Jane Mansbridge and Mark Warren, pp. 299–314. Oxford: Oxford University Press. Smith, Graham(2009): Democratic Innovations. Designing Institutions for Citizen Participation. Cambridge: Cambridge University Press. Steel, Daniel, Kristina Jenei, Michael Burgess, and Naseeb Bolduc (2020): Rethinking Representation and Diversity in Deliberative Minipub­ lics. Journal of Deliberative Democracy 16(1): pp. 46–57. doi:10.16997/ jdd.398. Stanford News Service(2017): Collaboration at Stanford leads to Mon­ golian parliament passing law on public opinion polling. https://news. stanford.edu/press-releases/2017/05/02/collaboration-st-opinion-polling/. Accessed: 18.08.2021. The Economist(2020): Citizens Assemblies are Increasingly Popular. https://www.economist.com/international/2020/09/19/citizens-assem­ blies-are-increasingly-popular. Accessed: 18.08.2021. UN Democracy Fund and newDemocracy Foundation(2019): En­ abling National Initiatives to Take Democracy Beyond Elections. https:// www.newdemocracy.com.au/wp-content/uploads/2018/10/newDemoc­ racy-UNDEF-Handbook.pdf. Accessed: 18.08.2021. Van Biezen, Ingrid, and Thomas Poguntke(2014): The decline of membership-based politics. Party Politics 20(2): pp. 205–216. doi:10.1177/1354068813519969. van der Does, Ramon, and Vincent Jacquet(2021): Small-Scale ­Deliberation and Mass Democracy: A Systematic Review of the Spill­ over Effects of Deliberative Minipublics. Political Studies(OnlineFirst). doi:10.1177/00323217211007278. Van Reybrouck, David(2016): Against Elections: The Case for Democ­ racy. London: Penguin Random House. Walker, Edward T., Michael McQuarrie, and Caroline W. Lee(2015): Rising Participation and Declining Democracy. In Democratizing Inequali­ ties. Dilemmas of the New Public Participation, Edited by Caroline W. Lee, Michael McQuarrie, and Edward T. Walker, pp. 3–23. New York/London: New York University Press. Warren, Mark E., and Hilary Pearse(2008): Introduction: democratic renewal and deliberative democracy. In Designing deliberative democ­ racy. The British Columbia Citizens’ Assembly, Edited by Hilary Pearse, and Mark E. Warren, pp. 1–19. Cambridge: Cambridge University Press. Warren, Mark, and John Gastil(2015): Can Deliberative Minipublics Address the Cognitive Challenges of Democratic Citizenship? The Journal of Politics 77(2): pp. 562–574. doi:10.1086/680078. Warren, Mark E.(2017): A Problem-Based Approach to Demo­ cratic Theory. American Political Science Review 111(1): pp. 39–53. doi:10.1017/S0003055416000605. Webb, Paul(2013): Who is willing to participate? Dissatisfied demo­ crats, stealth democrats and populists in the United Kingdom. Euro­ pean Journal of Political Research, 52(6): pp. 747–772. doi:10.1111/14756765.12021. Yeung, Peter(2020):‘It gave me hope in democracy’: how French cit­ izens are embracing people power. https://www.theguardian.com/ world/2020/nov/20/it-gave-me-hope-in-democracy-how-french-citizensare-embracing-people-power?utm_term=a58e00891ec399cc237052b­ 6d4612479&utm_campaign=ThisIsEurope&utm_source=esp&utm_medi­ um=Email&CMP=thisiseurope_email. Accessed: 18.08.2021. Young, I. M.(1996): Communication and the other: Beyond deliberative democracy. In Democracy and difference: Contesting the boundaries of the political, Edited by Seyla Benhabib, pp. 120–135. Princeton: Princeton University Press. 34

INTERVIEW PARTNERS Dr. Andrew Blick, Project Director of the CCUKDemocracy and Director of the Centre for British Politics and Government at King’s College Lon­ don(July, 2021) Prof. Dr. Brigitte Geißel, Director of the Research Center“Democratic Innovations” at Goethe-University, Frankfurt a. M.(July, 2021) Nicole Hartmann, Coordination Director for the Projekt Bürgerrat Klima (August, 2021) Felix Nasser, Coordination Director for the Projekt Bürgerrat Klima(Au­ gust, 2021) Claudine Nierth, Speaker of the Federal Board of Mehr Demokratie e. V. (July, 2021) Anna Quadflieg, from 2015 to 2020 Director of the Department“For­ eign Relations and Regional Development” in the Ministry of the German Language Community in Belgium(July, 2021) Anna Stuers, Permanent Secretary of the“Citizens’ Dialogue in Eastern Belgium”(August, 2021) Dr. Antoine Vergne, Co-Director of Missions Publiques(August, 2021) Iain Walker, Executive Director of the newDemocracy Foundation(July, 2021) INTERVIEW PARTNERS 35

imprint ABOUT THE AUTHORS IMPRINT Wolfgang Merkel is Director Emeritus of the Berlin Social Science Center, Professor Emeritus at the Humboldt Univer­ sity in Berlin, and a senior scholar at the Democracy Institute of the Central European University in Budapest. Filip Milacic is a senior researcher at the Friedrich-EbertStiftung’s office for“The Democracy of the Future” in Vien­ na and a visiting fellow at the Institute for Human Sciences in Vienna. Andreas Schäfer is a research associate in the Department of Political Sociology and Social Policy at the Department of Social Sciences at the Humboldt-Universität zu Berlin.. Publisher: FES Regional Office for international Cooperation Reichsratsstr. 13/5 A-1010 Vienna Contact: info@fes-vienna.org Responsibility for content: Johanna Lutz, Director, Democracy of the Future Contact/Orders: democracy.vienna@fes.de Design: pertext, Berlin| www.pertext.de The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung or of the organization for which the author works. Commercial use of media pub­ lished by the Friedrich-Ebert-Stiftung(FES) is not permitted without the written consent of the FES Publications by the Friedrich-Ebert-Stiftung may not be used for electioneering purposes. ISBN 978-3-96250-976-7 © 2021 www.fes.de/bibliothek/fes-publikationen 37