Hybrid types

Deliberation and voting can be integrated in various ways. Within the REDRESS project, we broadly distinguish five manifestations of hybrid democratic innovations. Within each of these forms, various design variations exist.

Together, these examples show how talking and voting can be interwoven within a single process. Citizens are given the opportunity to contribute at different moments and in different roles: as a discussion partner, as a co-designer of solutions, and as a voter. Different groups are involved, ideas can be tested along the way, rich input for policy and debate emerges, citizens become better informed, participants gain democratic experience, and the likelihood grows that policy will be both better considered and more broadly supported.

Digital platforms such as Decidim are examples of an interactive platform – a digital platform on which citizens do not merely respond to questions from policymakers but take on a proactive role. Such a platform integrates elements of voting and deliberation: citizens can, for example, propose statements that others respond to, or submit citizen initiatives. Policymakers can use this information to select themes, explore compromises, or find more broadly supported formulations. In some cases, expressions of support for citizen proposals can also lead directly to a formal referendum.

Digital platforms have great potential reach, since they are low-threshold for citizens in terms of time and location. By structurally bundling the votes and opinions of large groups, it becomes clear where opportunities lie for broadly supported policy, or the aggregation can help set the agenda.

There are also variants in which voting and deliberation are strongly linked, even if they do not take place simultaneously. One example is participatory budgeting, and specifically newer variants in which the voting takes place among the wider population rather than only among the citizens who engage in dialogue with one another. Examples include participatory budgeting in Maastricht, Antwerp, and Ghent. During such a process, citizens are invited to help think about how part of the city budget is spent. They take part in neighborhood meetings, where problems and ideas are discussed and themes are established. Project proposals are then submitted on those themes. Digital channels are used to collect proposals, which are subsequently discussed further. In a later phase, citizens vote on projects via an online platform, with information from the discussions taken into account. In some cases, extra weight is given to projects that emerge from intensive discussions.

The digital voting round makes it possible to reach thousands of citizens and to achieve a broadly legitimated basis for concrete investments. The neighborhood conversations ensure that projects become substantively stronger and better aligned with local concerns. Because the city indicates in advance that the winning projects will be implemented within the available budget, a clear link is also created between participation and policy. This can contribute to trust and support, as citizens see that their choices are actually implemented in the city.

A referendum is a clear and recognizable example of a vote in which a large group of citizens can take part. A prominent question here is how voters can be well informed about the referendum topic. In this hybrid model, a role is reserved for a randomly selected citizens’ panel of about 20 citizens that reviews the referendum proposal. The best-known example is the Citizens’ Initiative Review (CIR) in the U.S. state of Oregon. The citizens’ panel hears proponents and opponents, speaks with experts, and drafts a short, balanced summary in plain language that addresses the key arguments for and against the proposal. This text is sent out together with the official voting materials. Research shows that voters who read the summary have a better understanding of what the issue is about and vote more on substance than voters who rely solely on campaign information.

The role of the citizens’ panel here is more focused than that of the policy-forming citizens’ assemblies described below. After all, the referendum proposal has already been formulated – the panel focuses specifically on informing fellow citizens about its content. The potential of this model lies in better-informed voters. The visible diversity of the group gives voters confidence that no single interest group or party has set the tone, but rather citizens like themselves. The deliberative panel thus functions as a public information filter. This can strengthen both the quality of individual choices and trust in the process as a whole, thereby contributing to broader support for the outcome.

A third form involves supplementing a citizens’ assembly with a large-scale vote earlier in the process. The purpose of the vote can be to select a topic for the dialogue, to serve as an initial gauge of opinions, priorities, or values as a starting point for the discussions, or to test initial ideas or proposals from the discussions against a larger group.

One example is G1000 gatherings. In Belgium and the Netherlands, these are large gatherings in which hundreds to as many as a thousand citizens talk in rounds at tables over the course of a single day and formulate proposals. Prior to this, the wider population has been able to vote on the themes that will be addressed. Voting sometimes takes place digitally, so that the results are immediately visible and can be fed back to the room. G1000 processes above all demonstrate the potential for democratic experience among participants and for agenda-setting. Participants hear many different voices, practice making arguments, and learn to make choices together. In the end, a package of proposals remains that can serve as input for councils and executive boards.

Besides agenda-setting, input from the wider public can also be used to provide substantive input for a citizens’ assembly, usually made up of around 150 randomly selected citizens. This can be done, for example, through a Participatory Value Evaluation, in which citizens weigh policy positions within realistic constraints such as budget, time, and climate impact. Or through Pol.is, in which citizens respond to statements or initial proposals submitted by the citizens’ assembly, and may also add their own statements. Through data analysis, patterns of consensus in the votes become visible.

In the final variant, we find a combination of a formal referendum vote and a randomly selected citizens’ assembly, with the policy-forming role resting primarily with the latter. In most cases, the citizens’ assembly takes place first, and its outcome is then put to the wider population. The citizens’ assembly receives information, hears experts, weighs interests and arguments, and formulates one or more concrete proposals. The proposal or proposals are then put to all voters in a referendum. Ireland’s citizens’ assemblies on same-sex marriage and abortion are well-known examples. A diverse group of citizens developed proposals there for constitutional amendments, after which the Irish population approved those proposals in referenda. Other examples include Iceland, where a citizens’ assembly developed a proposal for a new constitution, and the Canadian province of British Columbia, where a citizens’ assembly advised on electoral system reform. In both cases, the proposal was subsequently put to the population in a referendum.

In this model, the potential for agenda-setting, substantive input, and broad support becomes clearly visible. The citizens’ assembly helps to sharpen the framing of the topic and translates what may be a polarized debate into a concrete compromise that voters can then decide on. The vote then serves as a broad test and confirmation of the proposals. In this way, public debate is structured and decisions are made that are both well considered and broadly supported.

The reverse order also occurs: the citizens’ assembly then takes place only after the referendum. The aim in that case is to translate a rejecting, general, or ambiguous outcome into concrete choices. The British citizens’ assembly on Brexit is an example of this. The assembly acknowledged the outcome of the referendum – leaving the EU – but examined which form of Brexit citizens found most desirable under various scenarios, and under what conditions. In the Netherlands, there are examples at the local level of citizens’ assemblies that developed concrete and broadly supported plans for waste policy or parking policy after a municipal council plan had failed to gain support in a referendum. For issues where there is support for policy change but it is unclear under what conditions and in what form, such a process can offer a solution. The outcome of the vote is then not seen as an endpoint but as a signal calling for further elaboration. By giving citizens a role in this, they can experience that their voice is taken seriously, and policy can become substantively richer and easier to explain to a divided public. In this way, an in-depth citizens’ assembly helps reconcile clashing expectations after a referendum, which in turn can contribute to greater understanding of the eventual choices made by politicians and government.