Democracy needs more than a firewall

Anyone talking about the crisis of democracy in Germany quickly ends up talking about the AfD. The debate revolves around firewalls and banning procedures, coalitions and the domestic intelligence service. That is understandable. But perhaps the crucial story begins much earlier. Not with a party that pushes against democratic boundaries, but with a society in which more and more people feel that political decisions are being made far removed from them.

The American political scientist Sheri Berman, Professor of Political Science at Barnard College, Columbia University, poses a question in the Journal of Democracy that Germany, too, should be asking itself: Why do democratic institutions become vulnerable enough in the first place for politicians who challenge their rules and norms to succeed? Her answer takes us far beyond parties and parliaments. Democracy may consist of political institutions. But whether it works depends just as much on the social conditions on which those institutions rest.

More than political institutions

This brings us back to an old insight from democratic research: a good constitution and regular elections are not enough. Democracy needs a society whose members are connected by a minimum degree of trust, belonging and shared responsibility.

Berman sees precisely these foundations as being under threat in the United States. Economic inequality and regional disparities have grown, while opportunities for social advancement have diminished. At the same time, people from different political camps increasingly inhabit different social worlds. Education, income, place of residence, media use, cultural attitudes and party preference increasingly coincide.

People dissatisfied with wow democracy works

That such dissatisfaction should not be equated with a rejection of democracy is shown by the Bertelsmann Stiftung’s latest Democracy Monitor. Fundamental support for liberal democracy remains remarkably high in Germany. Eighty-two per cent of respondents show particularly strong support for democratic values. For democratic elections, the figure is as high as 91 per cent. Between 2019 and 2025, support even increased slightly in six of the eight areas examined. It would therefore be wrong to conclude that German society is turning away from democracy itself.

People are much more critical, however, of how democracy actually works in practice. The gap is particularly striking when it comes to political parties: 77 per cent consider their democratic roles and principles important, but only 17 per cent rate their actual functioning particularly positively. For elections, the gap is considerably smaller: 91 per cent support their democratic principles, while 65 per cent are also satisfied with how they work in practice. The problem therefore seems to be less that people no longer want democracy than that they are dissatisfied with some of its political institutions.

When differences become camps

But dissatisfaction with political institutions is only part of the problem. There is also a social development to which Berman draws particular attention: political differences are increasingly becoming intertwined with social differences. What matters is not that people disagree. A democracy can withstand considerable political differences as long as its social dividing lines intersect. The conservative neighbour may be a trade union member, the left-wing colleague may belong to the same local shooting club, and people from cities and rural areas may share the same economic interests. Political opponents then repeatedly encounter one another as allies in other areas of life.

It becomes more dangerous when all these differences begin to coincide. People with different views no longer simply face one another; instead, social camps confront each other. A political opponent no longer appears merely as someone who thinks differently, but as someone who lives differently, uses different media, holds different values and belongs to a different world.

But what do elections do to us?

At this point, it is worth taking Berman’s argument a step further. Political institutions do not merely respond to social division. They can themselves help shape the way people perceive social differences.

Elections are indispensable to parliamentary democracy. They enable peaceful changes of power, hold governments accountable and, in principle, give every eligible voter an equal vote. The political scientist Armin Schäfer, Professor of Comparative Politics at Johannes Gutenberg University Mainz, whose research focuses particularly on social and political inequality, voter turnout and political representation, even points out that elections are comparatively less socially skewed than many other forms of participation. This is precisely why he regards declining and socially unequal voter turnout as a serious problem.

Politics as competition

But elections have another side. They organise politics as competition.

Parties have to make differences visible. They need distinguishable programmes, have to mobilise supporters and defeat opponents. Election campaigns compress complex political questions into alternatives: for or against, government or opposition, us or them. On election night, it is clear who has won and who has lost.

This is not an accidental flaw of electoral democracy. It is how it works. The crucial question, then, is this: What happens to a society when political participation takes place predominantly in this competitive form?

Parties unite – and sort

Historically, parties have performed an enormous democratic service. They bring interests together, draw up political programmes, recruit people for public office and make political decisions attributable to identifiable actors. A democracy without organised political will formation would face serious problems of its own.

Yet parties also sort social conflicts into political camps.

When the power to bind society weakens

That need not be harmful. Parties can even pacify conflicts by reconciling different interests within their own organisations. The traditional broad-based parties performed precisely this function: workers, employees, Catholics, business owners and different regional groups had to reach compromises with one another within large parties.

But this capacity to bind different parts of society together is weakening. The sociologist Steffen Mau, Professor of Macrosociology at Humboldt University of Berlin, points out that the role of parties as places where society forms its political will has greatly diminished. In eastern Germany in particular, many people have the impression that “those at the top somehow sort it all out among themselves.” The party is then no longer necessarily the link between society and the state. It can itself appear to be part of a political world to which many citizens have no access.

Low trust in political parties

The extent of this distance can be measured more precisely. According to the Democracy Monitor 2026, only 14 per cent of respondents say that political parties take sufficient account of citizens’ concerns. Confidence in their ability to solve political problems is also weak. In a Forsa survey from June 2026 cited by the Bertelsmann Stiftung, 55 per cent responded to the question of which party they trusted most to solve the problems of our time: none of them.

This reveals a remarkable contradiction: most people still regard political parties as a necessary part of democracy, yet they are deeply dissatisfied with how those parties actually perform. The crisis of trust in political parties is therefore not the same as a crisis of trust in democracy. This distinction is particularly important when considering citizens’ assemblies. The aim is not to replace a democracy that citizens have rejected. It is to complement a democracy that continues to enjoy broad support with procedures that can help address some of its perceived weaknesses.

Political equality does not end with the right to vote

The widespread sense that political parties do not adequately take citizens’ concerns into account is not merely an expression of a general loss of trust. Armin Schäfer’s research shows that political participation is in fact socially unequal. People with low incomes and lower levels of education vote less often. Electoral abstention is particularly pronounced in socially disadvantaged neighbourhoods. Those who live in an environment where many others also do not vote and view politics negatively are themselves more likely to be discouraged from political participation.

This can create a vicious circle. Certain social groups participate less. Their interests receive less political attention. This, in turn, reduces their expectation that participation can make a difference.

Whose wishes become policy?

Inequality does not, however, end with the question of who participates. It apparently extends to the question of whose political wishes are reflected in decisions. Political scientist Lea Elsässer, who researches social inequality and political representation at Johannes Gutenberg University Mainz, and political scientist Svenja Hense, whose research includes political inequality and political behaviour, together with Armin Schäfer examined how strongly the support of different social groups for specific political proposals was related to subsequent decisions in Germany. In their comprehensive study, they analysed 746 proposals for political change for which surveys designed to reflect the population as accurately as possible were available between 1980 and 2013.

The result runs counter to the democratic ideal of political equality. Decisions corresponded much more frequently to the political preferences of higher occupational and educational groups than to those of less privileged groups – regardless of whether the issues were economic or social and regardless of which parties were in government. The study therefore reveals a systematically unequal willingness on the part of politics to respond to the wishes of different social groups.

Decisions that favor high-income earners

The imbalance becomes particularly clear when the wishes of people on high and low incomes diverge. An earlier study by Elsässer, Hense and Schäfer examining 252 political issues between 1998 and 2015 found that the more strongly higher-income people supported a political proposal, the more likely it was to be adopted. No such influence was apparent for lower-income people – in some cases, proposals were even less likely to be adopted when they received particularly strong support from them.

So this is about more than who is more likely to vote. Social inequality may first manifest itself in unequal political participation, then in an unequal social composition of political institutions – and finally in whose political wishes have a greater chance of being reflected in actual decisions.

Who makes it into parliament at all?

Schäfer therefore speaks of a “loss of political equality.” Although every vote cast in an election carries the same weight, not all social groups are equally likely to make it to the ballot box, into parties, parliaments or other forms of political participation.

The composition of the 21st German Bundestag shows just how strong the selection process on the road to parliament is. Of its 630 members, only 204 are women – 32.4 per cent – although women make up slightly more than half of the population. The occupational imbalance is even more striking: 459 members – 72.9 per cent – are assigned to the occupational category of business organisation, law and administration. Among employees subject to social insurance contributions, by contrast, only around 20.5 per cent work in the comparable field of business organisation, accounting, law and administration.

Who is missing?

At the other end of the scale, the picture is almost reversed. Only 20 members of parliament come from occupations involving raw-material extraction, production and manufacturing – 3.2 per cent of the Bundestag. Among employees subject to social insurance contributions, however, around 19.7 per cent belong to this occupational category. Despite the caution required because the underlying populations are not identical, people from production occupations thus achieve only about one sixth of their share of the employment structure in the Bundestag.

The Bundestag is, of course, democratically authorised through elections to make decisions for the population. But it is not a miniature social image of Germany.

The invisible selection before participation

Nor can the problem simply be solved by offering more open opportunities for participation alongside elections. Schäfer’s research reveals an uncomfortable finding: many such opportunities are used disproportionately by people who are already well educated and politically active and can therefore actually increase political inequality.

Just how strong this self-selection can be was demonstrated, of all things, by the selection process for the first citizens’ assembly established by the Bundestag, “Nutrition in Transition.”

Those who volunteer

People were first contacted at random and asked whether they would in principle be willing to participate. Among those who subsequently volunteered, 71.5 per cent had a university or technical-college degree. At the time, only 26.3 per cent of the population belonged to this educational group.

Had the organisers simply selected those who volunteered, the result would therefore have been a massive social imbalance – even though the initial invitation had been random.

Sortition corrects the imbalance

This is precisely where stratified sortition came into play. From the pool of people willing to participate, members were subsequently drawn according to certain characteristics so that the citizens’ assembly would resemble the population as closely as possible. Among the members originally selected, 23 per cent had a low level of formal education, 61 per cent a medium level and only 16 per cent a high level. Even at the final in-person meeting, the proportions were still 23, 62 and 14 per cent.

It is a remarkable small democratic experiment: between invitation and actual participation, the same kind of social selection that shapes many conventional participation processes initially appeared. Sortition then corrected it.

Two different kinds of representation

This reveals an often underestimated difference between election and sortition.

Sortition does not make people a better reflection of society because randomly selected citizens somehow make better political judgements. Rather, it makes it possible to correct social imbalances that arise along the long path through political interest, party membership, internal party selection, candidacy and election.

Elections and sortition do different things

The Bundestag derives its democratic authority to represent citizens from elections. A citizens’ assembly, by contrast, can use sortition to reflect the social composition of the population far more closely.

These are not the same thing. And that is precisely why the two forms can complement each other.

Citizens’ assemblies have blind spots too

The Bundestag citizens’ assembly also demonstrates the limits of the process. People with an immigrant background were not included as a separate selection criterion. Of the 107 participants for whom corresponding information was available, only eleven – around ten per cent – had an immigrant background. In the population at the time, the figure was roughly 30 per cent. The scientific evaluation of the citizens’ assembly explicitly identifies this discrepancy as a weakness in its reflection of society.

Sortition alone, then, is not enough. What matters is how the selection is carried out.

Democracy must reach out to people

A simple random selection can reproduce social inequalities because people with higher levels of education and greater political interest are more likely to respond positively to an invitation. A good sortition process must therefore deliberately safeguard social diversity and, where necessary, make particular efforts to reach groups that are less likely to participate.

The sociologist Peter C. Dienel, Professor at the University of Wuppertal and inventor of the Planning Cell, a forerunner of today’s citizens’ assemblies, recognised this problem decades ago. He wanted neither to abolish party democracy nor simply to hand political decisions over to citizens’ initiatives or referendums.

Dienel’s democracy in need of supplementation

His starting point was instead that democracy needed to develop its procedures further. Party democracy, he argued, was “in need of supplementation.” His Planning Cell was intended to enable citizens to meet one another fairly, face to face, while remaining embedded in existing political decision-making processes.

Sortition changes what appears to be a minor but is in fact a fundamental question: Who gets into the room in the first place?

Who gets into the room?

In elections, people have to decide who should represent them. In parties, people first have to become active themselves. Those who want to attend come to public meetings. In citizens’ initiatives and referendums, those who feel particularly strongly about an issue become involved.

In a citizens’ assembly, the direction is reversed: democracy reaches out to the people.

A different logic of politics

The difference becomes even more fundamental once a citizens’ assembly begins its work.

A party has good reason to emphasise differences from its political opponents. A candidate has to defeat a rival. An opposition party may benefit when the government performs badly. Election campaigns reward sharp distinctions, recognisability and mobilisation.

No re-election, no opponent

These incentives do not apply to randomly selected members of a Citizens’ Assembly.

They do not have to be re-elected. They do not have to raise the profile of a party. They do not have to mobilise supporters or defeat a political opponent. After a few weeks or months, they returns to her ordinary life.

Different rules, different incentives

The political scientist Hélène Landemore, Professor of Political Science at Yale University, sees this as a fundamental advantage of representation by sortition. Today’s parliamentary democracy, she argues, in practice opens access to political power only to certain social groups. Her alternative model of “open democracy” is intended to bring ordinary citizens directly into the heart of political deliberation through sortition.

The point is not that randomly selected people are inherently more reasonable than politicians. The crucial point lies in the rules of the process: different rules create different incentives.

Suddenly, your opponent is sitting next to you

This brings citizens’ assemblies directly back to Berman’s original problem.

If modern societies suffer because different social groups encounter one another less and less, a well-composed citizens’ assembly deliberately creates such encounters. The university graduate from the big city sits next to the tradesman from a small town, the AfD voter next to the Green voter, the pensioner next to the student; people with and without an immigrant background deliberate on the same question.

From political camp to common task

In an election campaign, they may belong to opposing camps. In a citizens’ assembly, they have a common task.

They do not have to persuade one another. They do not have to abandon their fundamental political convictions. But they can no longer reduce the other person quite so easily to a party label. They hear why that person arrived at a particular view, what experiences lie behind it and which interests need to be taken into account.

Cross-cutting relationships

On a small scale, the citizens’ assembly thus creates precisely those cross-cutting social relationships whose disappearance Berman regards as dangerous.

Does it actually work?

The idea sounds almost too good to be true: put people with different political convictions around a table, let them talk to one another – and social divisions disappear.

It is not that simple.

Discussion can also deepen divisions

International research even shows that ordinary group discussions can intensify political differences. People reinforce one another’s views; like-minded participants can push one another towards more extreme positions.

With well-designed deliberative processes, however, the picture is different. What matters is a diverse composition, balanced information, differing expert opinions, sufficient time, independent facilitation and rules that encourage people to listen to one another.

Design matters

Under these conditions, studies find both a convergence of political positions and a reduction in hostility between opposing political groups.

That is a crucial distinction: conversation alone does not produce the effect. What matters is how democratic conversation is organised.

America in one room

This was demonstrated particularly strikingly in 2019 by “America in One Room.” A total of 523 American voters came together for several days. The group was composed to reflect the American electorate as closely as possible. Republicans, Democrats and independents deliberated together on immigration, the economy, healthcare, foreign policy and the environment.

Their political views were measured before and after the meeting and compared with those of a control group.

Less hostility

On numerous highly contentious issues, positions moved closer together. Particularly far-reaching initial positions frequently shifted towards the middle. But something else happened as well: dislike of supporters of the opposing political side also declined.

This touches directly on Berman’s diagnosis. Social camps apparently do not inevitably have to drift ever further apart. Under suitable conditions, people can retain their political differences while becoming less hostile towards the other side.

Camps become people

They do not necessarily change their political convictions. They experience something else: the people on the other side have reasons for their views.

“The Republicans,” “the left” or “the right” become actual people.

Democracy changes opinions differently

Here we see a distinction that extends beyond citizens’ assemblies.

Elections measure political views at a particular moment. A citizen chooses a party on the basis of the information and experience available to her. Democracy counts that decision.

From opinion to judgement

A citizens’ assembly asks a different question.

Not: What do you think now?

But: What judgement do you reach after you have studied the problem in depth, heard from different experts, met people directly affected and spoken with people who think very differently from you?

When your own voice counts

The American democratic theorist and former legislator Terrill Bouricius, who has worked for many years on sortition and the renewal of democratic institutions, considers this distinction fundamental. In a mass election, the influence of an individual vote is vanishingly small. There is therefore only a limited incentive for an individual to spend weeks studying difficult policy questions. In a randomly selected body, the situation changes. Participants know that their work and their judgement genuinely matter.

Political scientist Patrizia Nanz, who has worked for many years on citizen participation and the future of democracy and is now President of the European University Institute in Florence, and political and social scientist Claus Leggewie, Professor at Justus Liebig University Giessen and a long-standing researcher on democracy, society and political culture, have championed precisely this different form of democratic will formation with their proposal for a “Konsultative”: randomly selected citizens’ bodies permanently embedded as a complement to parliamentary democracy.

The experience of being able to make a difference

For Steffen Mau, this presents a particular opportunity in eastern Germany. After 1990, the western German party system was largely transferred to the East even though parties were less firmly rooted in society there. The hopes of the peaceful revolution for other forms of democratic participation played little role in constructing the political order of reunified Germany.

Mau therefore explicitly proposes citizens’ assemblies not as a replacement for parliamentary democracy but as a complement to it. He hopes they can provide “experiences of self-efficacy” and a “relative immunity to extremisms.”

From spectator to deliberator

His argument goes to the heart of the problem. For most people, democracy is currently experienced largely passively: politicians speak, citizens listen. Parties nominate candidates, citizens vote. Governments decide, citizens judge them at the next election.

In a Citizens’ Assembly, the role changes. For a limited period, citizens themselves become deliberators.

What this change of role can feel like from the participants’ point of view is illustrated by comments from the Citizens’ Assembly “Germany’s Role in the World.” “For the first time, I felt like a ‘mature’ citizen whose opinion was being asked,” one participant reported. Another described the political function of the process this way: “The Citizens’ Assembly can hold up a mirror to politics, spur it on and encourage it to provide answers.”

Even the dissatisfied can be reached

The importance of this is illustrated by a study of the Polish Citizens’ Assembly on energy poverty. Researchers wanted to know whether such processes reach only those who already have a positive attitude towards democracy.

After taking part, citizens reported greater political interest, more knowledge and a stronger sense of political self-efficacy. Remarkably, such effects were not confined to politically moderate or already satisfied people. Changes were also observed among participants who were politically dissatisfied or held more radical views.

Not political re-education

This is not evidence that citizens’ assemblies eliminate extremism. And it would be disastrous to treat them as a means of politically re-educating AfD voters.

But the finding does challenge a pessimistic assumption: people who deeply distrust politics are not necessarily beyond the reach of democratic experience.

Conflict does not disappear

Ireland also provides an example. Its Citizens’ Assembly deliberated in 2016 and 2017 on abortion law – an issue that had deeply divided Irish society for decades.

The citizens’ assembly did not resolve this moral conflict. Nor would that have been its task. People continued to hold different views.

Conflict without hostility

But the form of the dispute changed.

Participants heard from experts, people directly affected and representatives of different positions. Scientific studies of the deliberations show that the process enabled a multifaceted engagement with arguments.

A different form of disagreement

This may be one of the central democratic achievements of citizens’ assemblies: democracy does not need less conflict. It needs procedures in which conflict is not inevitably translated into hostility.

The limits of the Citizens’ Assembly

Research also counsels caution.

Within the French Climate Citizens’ Assembly, considerable mutual trust developed. Participants eventually even trusted working groups to which they themselves had not belonged.

Stronger inside the room than outside it

Outside the Citizens’ Assembly, the effect was much weaker.

This is one of the most important limitations of the process: citizens’ assemblies can demonstrably change the people who take part in them. It is much more difficult to extend that experience to millions of people who were not there.

One process is not enough

A citizens’ assembly of 100 or 160 people cannot by itself reconcile a divided society.

This is less an argument against citizens’ assemblies than against the idea that a single such process every few years could fundamentally transform a democracy.

Participation can also disappoint

Anyone who invites people to give up several weekends of their lives, listen to experts, deliberate with one another and draw up recommendations creates expectations.

If those recommendations subsequently disappear into a drawer, an experience of democratic self-efficacy can turn into its opposite: the experience that people were allowed to speak, but nobody listened.

Impact requires consequences

Studies from Ireland show that the effect of a citizens’ assembly on the wider population may depend substantially on what subsequently happens to its recommendations. If political decision-makers take the results seriously, public acceptance of the resulting decision can increase.

Particularly interestingly, this effect was strongest among people with low levels of political trust.

If you invite participation, you must respond

A simple democratic rule follows:

If you invite participation, you must respond.

A parliament does not have to adopt the recommendations of a Citizens’ Assembly. Elected representatives retain decision-making authority. But they should publicly explain which recommendations they are adopting, which they are changing and which they are rejecting – and why.

Silence is not an answer

A Citizens’ Assembly may deliberate. A parliament may decide differently. But it should not remain silent.

Politicians have also emphasised this connection between citizens’ assemblies and parliamentary democracy. “Democracy concerns us all. In the ‘Nutrition in Transition’ Citizens’ Assembly, democracy was lived, in a climate of openness, curiosity and the courage to engage in factual exchange,” said then-President of the Bundestag Bärbel Bas. FDP member of parliament Stephan Thomae even saw the Citizens’ Assembly as “a way of improving parliament and politics as such.” From this perspective, citizens’ assemblies do not stand in opposition to parliamentary democracy. They can add something to it.

What favours the AfD

All of this changes the way we look at the AfD’s success.

Its strength can of course not be explained solely by a crisis of democratic participation. Far-right, nationalist and xenophobic attitudes have causes of their own. Social alienation does not excuse a political choice.

The crisis began earlier

But it would be equally wrong to regard the democratic crisis solely as a consequence of the AfD’s rise.

A party whose narrative rests substantially on the opposition between “the people” and the “political class” encounters particularly favourable conditions when people genuinely feel that politics constitutes a closed social world.

Beneficiary and amplifier

The AfD is therefore both a beneficiary and an amplifier of a social development that began long before its rise.

Citizens’ assemblies have a remarkable advantage in confronting this narrative. A parliament can be accused of overrepresenting certain occupational and educational groups. Parties can be attacked as organisations of professional politicians. But it is considerably harder to dismiss a cross-section of the population selected by age, gender, education, place of residence and other characteristics as “those at the top.”

From the heart of society

Mau sees in this a certain resilience of citizens’ assemblies against the accusation that politics is dominated by a detached ruling class. Their participants visibly come from the heart of society.

Not elections versus sortition

None of this should lead to a false choice: bad elections on one side, good citizens’ assemblies on the other.

Elections achieve something a Citizens’ Assembly cannot. They allow the entire eligible electorate simultaneously to confirm or remove governments. They enable decisions about the broad political direction of a country. They make responsibility visible. And despite all social differences, as Armin Schäfer shows, they are less socially skewed than many voluntary forms of participation.

What citizens’ assemblies can add

Citizens’ assemblies, in turn, can achieve something elections can hardly do: bring together people with very different life experiences, give them time to learn, remove them from the immediate competition between parties and allow considered judgements to emerge from spontaneous opinions.

It may be precisely the difference between the two processes that constitutes their strength.

Two different questions

Elections answer the question: Who should govern?

Citizens’ assemblies answer another: What would different citizens recommend if they had time to study a problem thoroughly together?

A vibrant democracy may need both.

From individual citizens’ assemblies to a democratic institution

This leads the research to a further conclusion.

If the strongest effects occur among those who themselves take part in citizens’ assemblies, then in the long run it is not enough to convene a hundred randomly selected citizens every few years.

A lasting experience rather than an exception

The real democratic effect might emerge if citizens’ assemblies became a permanent part of the political system: at federal, state and local level. When sortition becomes an institution.

Over years and decades, it would then not be hundreds but tens of thousands of people who had experienced taking part in political deliberation themselves. Families, workplaces, associations and neighbourhoods would include people who had once been selected by lot.

When citizens’ assemblies become normal

Citizens’ assemblies would then no longer be exceptional participation projects occasionally covered by politicians and the media. They would become a normal democratic experience.

This is precisely the direction in which Patrizia Nanz and Claus Leggewie point with their proposal for a permanently established “Konsultative.” Hélène Landemore likewise sees sortition not as an occasional supplement but as part of a fundamentally more open democracy.

The firewall is not enough

Berman’s analysis therefore leads to an uncomfortable conclusion for Germany’s approach to the AfD. A democracy cannot defend itself solely by fending off its opponents.

Firewalls can limit political cooperation. Courts protect the constitution. The domestic intelligence service monitors extremist activities. All of this may be necessary. None of it, however, answers the question of why people come to feel that democratic politics has less and less to do with them.

The side effects of political life

Still less does it answer the question of whether our democratic procedures produce unintended side effects: social selection, professionalised politics, short time horizons, the formation of camps and the pressure to translate social differences repeatedly into winners and losers.

Decades ago, Peter C. Dienel therefore questioned not democracy itself but its procedures. Party democracy, he argued, was “in need of supplementation.”

Democracy is more than voting

Hélène Landemore takes the challenge further today: democracy should not be equated with elections. Democratic representation can also arise through sortition – giving ordinary people access to political power that electoral democracies grant only to a small minority.

Steffen Mau reaches a similar conclusion from his analysis of contemporary Germany. He describes himself as a “friend of citizens’ assemblies as a complement to party democracy”: places where people can talk to one another about concrete issues and reach compromises.

Perhaps the problem runs deeper

The real question, then, is not whether citizens’ assemblies weaken the AfD.

It is what kind of democracy a society produces when its citizens experience politics almost exclusively as spectators of a competition between parties.

What electoral democracy is missing

Elections require decisions. Parties need differences. Election campaigns need sharp distinctions. Democracy needs all of this – but perhaps it needs something else as well.

Places where nobody has to win the next election.

Places for deliberating together

Places where a randomly selected shop assistant has the same political weight as the professor, the business owner or the party official.

Places where people do not appear as groups of voters, target groups or political camps, but first and foremost as citizens.

Where the opponent does not have to be defeated

Places where a political opponent does not have to be defeated because, in the end, everyone is supposed to work together on a recommendation.

Citizens’ assemblies are no miracle cure for social division. They cannot eliminate inequality, replace an effective public administration or prevent bad policy. If their recommendations subsequently have no consequences, they can even create new mistrust.

But they change the rules of the game

But they change the rules of the game.

Competitors become deliberators. Spectators become participants. A single vote among millions becomes a person who is trusted to form a judgement.

What the research brings together

Sheri Berman describes a society whose groups are becoming increasingly alien to one another. Armin Schäfer shows a democracy in which social inequality can become political inequality. Steffen Mau describes people who lack the experience of being able to make a political difference. Dienel, Nanz and Leggewie, Landemore and Bouricius all point, in different ways, to the fact that elections and parties are not the only conceivable means of forming democratic political will.

Research now adds something crucial: under suitable conditions, randomly selected and deliberative processes can in fact bring about some of the changes these arguments would lead us to expect – greater political self-efficacy, greater understanding of other viewpoints and less hostility between political camps.

No miracle cure – but more than an experiment

Not always. Not for everyone. And certainly not automatically throughout society.

But often enough to take seriously the question of whether citizens’ assemblies should continue to remain merely an occasional addition.

More trust in citizens

Perhaps the more important answer to the crisis of democracy, then, is not to explain the existing democracy to citizens ever more effectively.

A democracy suffering from a loss of trust should not merely explain to people why they ought to trust it.

It could place more trust in them.

Note: The image accompanying this article was created using AI.