Aristotle meets Rousseau: recent citizens assemblies (G1000s) in the Netherlands

Geerten Boogard and Harmen Binnema

Since parliamentary democracy has come under fire in many western European democracies, two ancient concepts from constitutional theory reentered the stage: sortition and the citizens assembly. David van Reybrouck took the Aristotelian claim that pure democracy requires a form of lottery and combined it with the dream of Rousseau that the ultimate sovereignty of the people lies in their own assembly. Although neither sortition nor citizens assembles are really new, Van Reybrouck made his claim so eloquently that he inspired many fellow citizens, both in Belgium and elsewhere.

A highly visible result of Van Reybrouck’s work are the G1000s. Van Reybrouck (and others) organized the first one in 2011 in Brussels, for which they randomly selected 1000 Belgians to deliberate for one day on major policy issues. This Belgian success inspired other countries, among which northern neighbors the Netherlands. Since the start early 2014 in the town of Amersfoort, a substantial number of G1000s have been organized at the local level. We have been in the privileged position to study a couple of those citizens assemblies in very close detail.

In this contribution we will reflect on two aspects of the G1000 which seem to be crucial in its claim to renew democracy. First, the diversity of the participants, which arguably should be enhanced as a result of sortition. Second, the agenda setting and policy influence that a G1000 is expected to have, being an alternative for representative democracy. We conclude by providing some suggestions to remedy the shortcomings of the current design of the G1000.


For the large majority of G1000s held in the Netherlands, the participants were randomly selected, by lot. A sample was drawn from the municipal personal records database, and those citizens included in the sample would receive a letter from the local organizers – in some cases signed by the mayor – to invite them to participate in a G1000. The letter explained that the purpose of a G1000 is to think about the future of the municipality for the next decade: where do you want your city to be in 2020 or 2030? Compared to jury duty, there is no obligation to participate. Moreover, participants do not receive a financial compensation. A G1000 normally lasts about six to seven hours, including coffee and lunch break (Saturday morning and afternoon). In other words, it appeals to a particular kind of active citizenship, a willingness to engage with the community, of which the reward or benefit is unclear.

Theoretically speaking, the diversity of the participants should increase as a result of random selection. Yet, what we found in all G1000s was that only 5 to 10% of those invited actually showed up. Therefore, the difference in terms of diversity with procedures of  interactive policy making or public hearings, is negligible. To be sure, the number of participants (250 to 600) clearly exceeds that which we see in other forms of citizen participation. If we look at the demographic and socio-economic background of the participants, however, it is much more of the same. The lower educated and the young are underrepresented, and the same holds for ethnic minorities. Jacquet (2017) names three important reasons for non-participation: concentration on the private sphere, conflict of schedule and the lack of impact on the political system. In addition, G1000 participants often are active citizens who are more likely to vote in local or national elections. With this high degree of absenteeism, sortition eventually turns into self-selection.

Citizens’ assembly

The G1000 is considered an important alternative to the traditional structures of representative democracy, as it provides citizens with a direct instead of an indirect say in local politics. In addition, it is argued that the deliberative nature of the G1000s, initiated by a Dutch Platform called, stimulates citizens to look for what they share (common ground) instead of what separates them. A crucial role is ascribed to the Agenda for the City, consisting of the ten proposals that have been developed during the day and voted upon in a final round. First, since they reflect the ideas of the participants themselves, these proposals are expected to be taken up by groups of citizens and put into practice. Second, they are expected to have an impact on the local policy agenda, to set the priorities of local politicians.

Yet, we found that both of these ambitions hardly materialized. Several thematic groups or project groups that were formed to work on one of the proposals from the Agenda for the City, did not last long, for all kinds of reasons (mostly practical). Moreover, the proposals did not find their way to local politics. Although politicians took notice of the Agenda, e.g. by participating in a G1000 themselves, they not turn this into political action. Some referred to a conflict with their own party platform, others argued that many of the proposals were not very new or innovative, or already part of existing local policy.

A number of factors may explain this lack of policy impact. First, the difference between what we call the citizen summit and the citizen council. While the first part of the G1000 (summit) focuses on collecting dreams, desires, ideas, the second part (council) turns the G1000 into a decision making body, which challenges the local council as an alternative representation of citizens. This can also be seen in the dual expectations that participants expressed: some were there to meet like-minded with whom they could organize collective action, others considered it an opportunity to tell politicians what they had to do and how they wanted local politics to change. The second explanation is that the G1000 in its current (Dutch) format does not involve the role of experts who could indicate what kind of proposals are more viable, which options have been tried before or what the costs and benefits of certain measures could be. In some cases, this led to proposals being developed (and even making it to the top-10) that already existed. 

Bringing the G1000 forward

Where should we go then? Given the considerable amount of what we like to call ‘democratic energy’ a G1000 attracts and creates, we will look for ways to optimize a G1000 in the system of the local government. Instead of just concluding that the results so far did not exactly live up to the original ambition of Van Reybrouck and others, we prefer to improve the design of the G1000, whilts maintaining the aspects of sortition and citizens assembly.

Starting with sortition, we argue that diversity can be increased by providing some compensation for the participants. A recent experience in Utrecht showed that this at least increased turnout among students. Next to adding a financial stimulus, we suggest that the sense of urgency of a G1000 should be increased and be more directly related to important local issues. In various municipalities, the issue of housing for asylum seekers and refugees, brought hundreds of citizens to hearings of debates in the city hall, church or sports centre. At the same time, G1000 organizers were struggling to incite citizens to participate on the very broad noition of the ‘agenda for the city’.  

Second, we think that diversity can be increased by lowering the thresholds for participation. Some people are willing to participate, but hesitate to go on their own. In that case we can think of passing on or sharing the invitation or providing some kind of ‘wild card’ to take a friend, relative or neighbour with you. Other people do not answer the call of the G1000 because they expect the summit to be very much language based, which is indeed what deliberation and dialogue entails. But there are other forms of communication than just language which could be experimented with, like music, painting, constructing. In addition, least, there are a number of practical issues that have to do with location and timing of the G1000. Is the summit somewhere in the neighbourhood or is it in a remote building? Does the G1000 have to be on a Saturday, which is also a day for sports activities, family visits or shopping?  

Third, we argue that the idea of sortition should not be applied too dogmatic. As long as participation is not the same as jury duty, some level of self-selection will always be there. Knowing that in the optimal situation some 600 citizens showed up after thousands of invitation letters, there is a need to complement sortition with other means of selection. We know from our research which social groups (youngsters, ethnic minorities) are underrepresented. First, one could increase the number of invitations sent to these groups and second, they can be approached through existing networks and organisations. This is what was done in the last weeks prior to the G1000 in Belgium when the organizers noted a lack of diversity in the applications.

Regarding the aspect of citizen assembly, we argue that there are several ways of improving its place within the existing system. First, a choice needs to be made between the G1000 as a summit or as council. That is, either focusing on an open agenda, follow-up on themes by participants themselves and a limited role for politicians, or focusing on influence which implies reordering the priorities of local politics and connecting with urgent local issues. Second, to integrate the G1000 in the structure of political parties. One national political party already used a G1000 format in its process of writing the manifesto, while a municipality is designing such a meeting to provide ideas for all the local manifestos at once. Political influence is much more guaranteed in this way, than it is when 1000 citizens present their Agenda’s the day after the day of the election.


Edit this page