Setting the agenda for climate assemblies. Trade-offs and guiding principles

ABSTRACT

Citizens’ assemblies on climate change are increasingly popular to support democratic decision-making. Such Climate Assemblies (CAs) convene representative groups of citizens formulating policy proposals after hearing experts and deliberating intensely. CAs may help addressing climate policy issues more effectively partly because their members need not worry about re-election. CAs’ effectiveness depends on their design such as the issues chosen (or not chosen) for deliberation. Agenda-setters exert substantial power by selecting certain issues and by choosing framings that benefit some solutions over others. In this paper I ask: What characterizes agendas that are suitable and legitimate for deliberation in CAs? The aim is to support practitioners in making informed agenda choices for CAs by providing a list of ten widely accepted guiding principles based on expert interviews, policy documents, and information gathered from the Knowledge Network on Climate Assemblies (KNOCA). The paper systematically discusses trade-offs of various agenda choices in the light of different CA rationales. Results show that those with system-supporting rationales tend to favour narrower agendas tailored to political demands aiming to increase immediate policy impact; those with system-disrupting rationales prefer more open agendas allowing citizens to challenge existing political practices and worldviews. Results support earlier arguments that distinctions of entire deliberative processes in either top-down or bottom-up are too simplistic and that a tool-box approach is more useful. Insights appear relevant for debates of deliberative minipublics more generally. Future research should investigate whom to involve in setting CA agendas and with how much power.

Key policy insights

• Effective agenda design hinges on rationales on how to achieve assembly objectives which depend on authorities’ ambition for climate action.

• Given high ambition, system-supportive rationales aiming for policy impact favour narrower agendas tailored to demands of the policy process but risk low transformativeness.

• Given low ambition, system-disruptive rationales aiming to challenge established practices and worldviews favour more open agendas but risk low impact if assemblies are not politically embedded or able to mobilize opposition groups.

• Assembly designs are seldom purely supportive or disruptive but often hybrid.

• Agenda-setting has many dimensions allowing for productive combinations of disruptive and supportive elements tailored to contexts.

KEYWORDS:

• Deliberative minipublics
• participation
• climate change
• agenda-setting
• design
• impact

1. Introduction¹

Citizens’ assemblies on climate change are becoming increasingly popular as a tool to support democratic deliberation and decision-making. In the past five years, such Climate Assemblies (CAs) were convened at the national level in European countries like Ireland, France, the UK, Scotland, Spain, Denmark, Finland, Austria, Germany, and Luxembourg.² CAs convene a representatively selected group of lay-citizens to deliberate on possible solutions and pass on their proposals to policymakers. CAs consist of sixty to two-hundred participants from diverse backgrounds that receive multidisciplinary scientific input and may hear affected groups before intensely deliberating in small groups across multiple sessions spreading several weeks. Many hope that CAs can drive more ambitious, just, and effective climate policy (e.g. Dryzek & Niemeyer, 2019; Ejsing et al., 2023; Niemeyer, 2013; Willis et al., 2022), partly because CA members do not have to worry about re-election or their political career when it comes to taking a position on hot issues (Willis, 2020). In particular, they do not have to serve a special constituency which may expect specific policy priorities that are not in line with climate protection. By signalling public support CAs can provide policymakers with a ‘social mandate’ offering opportunities to legitimize policy change (Howarth et al., 2020). In an exemplary case in Ireland, citizens’ assemblies were able to facilitate collective decision-making on thorny issues like abortion and same-sex marriage (e.g. Farrell et al., 2019).

While most CAs share common features like sortition and high-quality deliberation, their design differs substantially, for example in the ways they are connected to conventional decision-making structures (Boswell et al., 2022; Smith, 2023). Effective design is crucial if CAs are to play a productively embedded role in democratic systems (Bussu et al., 2022) and drive ambitious climate policy (Wells et al., 2021). One among many important design decisions is setting the agenda for a CA. By selecting which issues (not) to discuss, and how to frame these issues, agenda-setters exert strong influence over the nature of deliberations and the political impacts of CAs (Elstub et al., 2021; Schattschneider, 1977).

Practitioners, such as commissioners or process designers, face difficult trade-offs when making agenda decisions. For example, is it more important to constrain the agenda tightly in order to enable structured deliberations that are more likely to result in climate policy proposals policymakers will take up, or to leave the agenda open, empowering citizens to discuss solutions that may challenge dominant unsustainable practices and agendas? Disagreement and confusion over the ‘right’ design choices largely stem from different objectives and rationales CA practitioners and academics adopt (see Hammond, 2020).

This paper addresses these confusions and enables more informed CA design choices that may enhance the effectiveness of future CAs. Its aim is to support practitioners in choosing suitable and legitimate agendas for CAs by providing a list of ten widely accepted guiding principles, and systematically discussing trade-offs of various agenda choices in the light of different rationales and contexts. The insights are based on expert interviews, and information gathered from the Knowledge Network on Climate Assemblies (KNOCA) as well as policy documents related to CAs in Europe. The focus is on national-level CAs, although many insights may be relevant for deliberative minipublics more generally, and at other levels of political authority.

In the next section, I cover theoretical backgrounds including different contexts and rationales related to CAs. This is followed by details of the conducted expert interviews, in section 3. In section 4, I present my results in the form of ten guiding principles and discuss the trade-offs of related agenda choices. I close by discussing the findings and pointing out avenues for further research.

2. Theoretical backgrounds

2.1. Climate assemblies

CAs are a special form of deliberative minipublic. Deliberative minipublics are participatory processes that are characterized by random selection of lay-participants, provision of balanced information, and facilitated small-group deliberations which culminate in a collective output, such as policy proposals (Smith & Setälä, 2018). Participants are typically encouraged to adhere to communicative norms of deliberation, such as equality, justification of arguments, or respect of and openness to others and their views (e.g. Gutmann & Thompson, 2009; Thompson, 2008). Citizens assemblies are a special form of minipublic characterized by a relatively large number of participants and long deliberation times (Smith & Setälä, 2018). CAs are citizens’ assemblies on climate-related issues (Boswell et al., 2022).

Several scholars argue that CAs can foster progressive climate policy, for example because deliberation in CAs fosters long-term thinking and increases willingness to pay for climate measures (MacKenzie & Caluwaerts, 2021); CA members do not have to worry about re-election; CAs systematically integrate expert knowledge; CAs limit the influence of fossil fuel lobbyists (Willis et al., 2022); or CAs alter public discourse (Dryzek & Niemeyer, 2019). Moreover, past CAs have repeatedly proposed measures that are more ambitious than current government policies (Hoffmann, 2023; Lage et al., 2023; Weber, 2023), such as a ban of domestic air travel in France if train alternatives exist.³

The effects of CAs on policymaking depend on their design. While CAs share many design features like the random selection of participants or intense deliberation, they also differ considerably in the way they are connected to conventional decision-making structures (Boswell et al., 2022). In Scotland, for example, the CA was commissioned by government who were legally required to respond to the final report within six months. In Germany, the CA was initiated by a civil-society actor with no formal commitments to respond by political actors.⁴ The agenda of CAs is one important design feature that has received little attention.

2.2. Climate assemblies and agenda-setting

Agenda-setting, according to Barbara Sinclair, is ‘the process through which issues attain the status of being seriously debated by politically relevant actors’ (1986, p. 35). Roger Cobb and Marc Ross note, ‘[a]genda conflicts are not just about what issues government chooses to act on; they are also about competing interpretations of political problems and the alternative worldviews that underlie them’ (1997, pp. 3–4). Hence, agenda-setters exert substantial power over public policymaking (Schattschneider, 1977).

Deliberative minipublics like CAs can have an agenda-setting function in the cycle of public policy processes (see Dahl, 2008/1989; Gastil & Richards, 2013 for ideas), e.g. by selecting issues to be discussed by elected officials as is practiced in East-Belgium (Niessen & Reuchamps, 2019). However, this is not what I refer to as agenda-setting in this paper.

Here, agenda-setting for CAs refers to setting the boundaries and guidelines for deliberations within the deliberative minipublic. Agenda-setting includes the choice of a general issue for deliberation, the framing of the process including the formulation of a remit, and the selection of sub-themes (or even proposals) to discuss or not discuss (i.e. constraining deliberation) (also see Barisione, 2012). In this study, I focus on the content and not the process of agenda-setting. That is, I do not address who should have the authority to initiate CAs, and set agendas, or how such processes should be designed (see Brancaforte & Pfeffer, 2022; Courant, 2022; Lang, 2008; and the discussion on this). I only address what agenda-setters, regardless of who they are, should generally consider.

Only few studies have addressed the agenda-setting for deliberative minipublics (Barisione, 2012; Blue, 2015; Bua, 2012; Courant, 2022; Lang, 2008; Parkinson, 2006). Parkinson (2006, Ch. 6) found that tightly set agendas following a bureaucratic-instrumental rationale to ‘solving a delineated problem’ can imply preference assumptions that many deliberating citizens do not share. Parkinson (2006) argues, the more open the agenda the more legitimate the process. Political elites can also exert strong influence over minipublic agendas, up to the point of co-option. That means, there have been incidences where participatory processes were used to legitimize decisions that had in fact already been taken (as summarized by Bua, 2012). Elite control over agendas has also led to the adoption of technocratic and economic framings in minipublics (on climate change), omitting various dimensions of justice (Barisione, 2012; Blue, 2015). On the other hand, constraining agendas in minipublics may be necessary to enable high quality deliberation and arrive at decisions (Lang, 2008). Parkinson (2006, Ch. 6) noted, the more open the process, the more vague are the outcomes, and the less likely practical solutions and policy impacts. Finally, Elstub et al. (2021) found that a large scope – a high number of issues to be covered by the minipublic – has reduced the policy impact of the UK CA, because it led public officials to doubt whether all participants had sufficient time to work on each issue. They also argued that it may be more difficult to hold policymakers accountable regarding their response to CAs, if the number of recommendations is too high. This would make it easier for policymakers to cherry-pick recommendations, meaning they select only those that are in line with their preferences (Font et al., 2018).

A number of practical guides exist to support practitioners in setting the agenda for minipublics (Carson, 2018; Pfeffer, 2022; Rourke, 2014). These practical guides are useful but do not address trade-offs in a systematic way, and tend to carry the either system-disruptive or system-supportive perspective of the author (see next section). This short recap foreshadows that practitioners face difficult trade-offs when making agenda decisions.

2.3. Climate assembly contexts and rationales

When discussing CAs, individuals hold divergent rationales and contexts in mind (Averchenkova & Ghilan, 2023; Bussu & Fleuß, 2023; Hammond, 2020; Rangoni et al., 2021; Sandover et al., 2021). This can lead to disagreements or misunderstandings over the roles CAs ought to play or how to design them effectively. I will introduce a number of common contexts and rationales to help limit misunderstandings. They will serve as the analytical framework to which I refer when discussing the trade-offs of agenda decisions.

The climate political context refers to widely shared informed beliefs about the main barriers to climate action (including both mitigation and adaptation). These barriers may include a lack of government motivation, internal government conflicts, powerful lobby interests impeding climate action, government hesitance due to fear of public resistance and voter loss, entrenched worldviews and paradigms conflicting with climate action, or insufficient governmental capacities (knowledge, time, money) (also see Jordan et al., 2022; Rickards et al., 2014). This list is exemplary and by no means exhaustive, and barriers may coexist simultaneously.

One can distinguish between two ideal-typical rationales of CAs with regards to climate change (Hammond, 2020).⁵ A system-supporting rationale primarily aims to support policymakers in addressing climate change, assuming a context where influential policymakers are generally willing to act. This rationale highlights the importance of cooperation and working toward agreement or even consensus to generate collective power to act (Partzsch, 2017), and the need to design processes aligning with political and administrative requirements to increase the uptake of proposals. A system-disrupting rationale, on the other hand, underscores the need for conflict, public attention, and pressure, along with more open deliberations that challenge dominant practices and paradigms in policymaking. This rationale often assumes a lack of willingness to act among policymakers and emphasizes needs for fundamental systemic transitions that would not occur through system-supporting approaches. The system-disruptive rationale often goes along with an emphasis of emancipatory objectives seeking to empower citizens, particularly marginalized groups, in democratic decision-making. It is crucial to recognize that these are ideal-typical rationales underlying individuals’ attitudes towards CAs and their design, and not descriptions of real CA cases.

Still, these rationales are often used to make rather black-and-white distinctions between deliberative processes, labelling them either top-down (i.e. supportive) or bottom-up (i.e. disruptive). When analyzing real cases, this distinction does not hold because many deliberative processes contain both system-supportive and system-disruptive elements (Bussu & Fleuß, 2023). The French CA, for example, was initiated top-down on behalf of the President but simultaneously gave members high degrees of freedom in developing their own ideas.⁶ This facilitated the adoption of disruptive proposals such as changing the constitution to make ecocide a crime. Hence, the distinction of system-supportive and system-disruptive rationales is useful in analyzing deliberative processes but only if applied to single elements as opposed to the entire process. In doing so, the ideal-type distinctions can "be used as a ‘toolbox’ that allows for flexible design and combination in democratic processes" (Bussu & Fleuß, 2023, pp. 150–151). In this spirit, I will use it throughout the paper to highlight trade-offs of various agenda choices.

3. Methods

I conducted 14 semi-structured expert interviews between February 10, 2022 and April 22, 2022. Experts were recruited from the Knowledge Network on Climate Assemblies (KNOCA). Expertise was defined as previous experience with organizing, commissioning, studying, or advising CAs. I selected the experts as to ensure diversity regarding their role(s), their location, their gender, and their rationales.

The interviewed experts have served as commissioners, public officials, process advisors, academic researchers, organizers and facilitators, and advisory activists.⁷ They have been involved in CAs in at least 14 different countries and at supra-national levels. 12 of 14 experts live in western Europe, and one each in Australia and the United States. 5 experts were female. 2 experts pre-dominantly adopted system-disrupting perspectives.

The interviews were semi-structured with the objective of developing a checklist of aspects for commissioners to consider when setting the agenda for CAs. First, I posed an open question asking experts for such aspects. Subsequently, I asked experts whether they agreed or disagreed with a list of further aspects and asked them to explain their selection. I previously collected these aspects together with two political science students who are members of a German organization advocating for and advising policymakers on CAs. This list of aspects and the interview guide were reviewed by the KNOCA chair. To surface underlying assumptions and theories, and to understand diverging answers, I asked experts to explain their answers, or invited them to respond to conflicting answers given by their anonymized peers. For validation, all experts had the opportunity to comment on the preliminary results; and the results were presented to scientific and non-scientific peers at the KNOCA annual meeting 2022, and the 2022 annual conference of the European Consortium for Political Research (ECPR).

The interview recordings were transcribed and analyzed in MAXQDA following established procedures of qualitative content analysis (Mayring, 2014). The data was coded according to the aspects to consider when setting the agenda, experts’ support or opposition toward the aspect, and their explanation and discussion points. Similar explanation and discussion points were categorized inductively. I developed the guiding principles presented in this paper by recursively incorporating the expert responses. Toward the last interviews, theoretical saturation was reached in the sense that experts did not provide novel themes or meanings, i.e. explanations or discussion points, and agreed with conclusions I proposed (Glaser & Strauss, 1967; Hennink et al., 2017).

The results were supplemented by information gathered from KNOCA⁸ – a European Network gathering knowledge and facilitating exchange on CAs – as well as policy documents related to CAs in Europe. KNOCA provides information on all national-level European CAs, hosts learning calls, and publishes research reports. I sought this information to validate and illustrate arguments made by interviewed experts. This was facilitated by the author being a contributing member of KNOCA and possessing extensive knowledge of their published information and CAs in Europe previous to the research. I point to supplementary information in footnotes. All online resources were last checked September 22, 2023.

4. Results: guiding principles for setting the agenda of climate assemblies

Table 1 summarizes my findings in the form of guiding principles with trade-offs presented as commentary by system-supporting and disrupting rationales (see last paragraph in theory section). In the following sections, I explain and discuss each principle and the commentary more elaborately.

Table 1. Guiding principles for agenda-setting in climate assemblies.

Guiding principle	Supporting view	Disrupting view
Scope. All participants have sufficient time to develop recommendations, understand consequences, and provide justifications.	Narrower scope increases policy impact	Wider scope allows space for challenging dominant agendas
Authority. The target authority has sufficient power to act on proposals of the selected agenda.	Necessary to allow for policy impact	Constrains deliberation to powerholders’ agendas
Societal relevance. The issue is important for citizens.	Less important for policy impact through bureaucratic pathways	Important to gain public attention and spark discourse, and to empower citizens
Political relevance. Policymakers see a need for change on the issue.	Increases policy impact	Constrains deliberation to powerholders’ agendas
Receptiveness. Policymakers welcome citizen input on the selected agenda.	Increases policy impact Uptake rules and interactions increase receptiveness	Policymakers tend to favour low risk issues Public pressure increases receptiveness
Timing. There is an opportunity to affect change.	Alignment with policy processes increases policy impact	Minipublic as window opening event
Dilemmas. Clear trade-offs must be made.	(No difference)
Legitimacy. The agenda is seen as legitimate by most groups.	Aiming for consensus is democratically desirable Perceived procedural legitimacy increases policy impact	Conflict can spark public attention and provoke critical discourse Need for consensus prevents more transformational agendas
Openness. Citizens are not unjustly constrained by the agenda.	Constrained agenda necessary for useful proposals	Citizen empowerment through more openness
Resource efficiency. Societal benefits of citizen deliberation on the selected agenda outweigh invested resources.	(No difference)

Source: Adapted and extended from Pfeffer (2022).

4.1. Scope

All participants have sufficient time to develop proposals, understand consequences, and provide justifications.

Here, the scope is understood as the (inverted) time a participant is given to deal with a proposal. In other words: a larger scope means less time. The scope is a function of the number of issues and proposals to be discussed during the process, the complexity or knowledge intensity of these issues and proposals, the overall deliberation time, and the structure of deliberations such as the division of labour between participants. Understanding scope in terms of time per proposal highlights that it can be subject to design. For example, a CA covering multiple sectors may still ensure sufficient deliberation time per proposal if the number of proposals is limited from the outset, as was practiced by the German assembly on nutrition.

Finding the ‘right’ scope of a CA is a frequently discussed issue (e.g. Elstub et al., 2021). To date, most CAs have had wider scopes. They often developed up to or over 100 proposals on issues spanning from renewable energy to food and agriculture or housing simultaneously (e.g. France, Scotland, UK, Austria, Germany).⁹ The Finish climate assembly provides an example of a narrower scope where participants were asked to appraise 14 policy proposals developed by the government (Kulha et al., 2022).

Experts interviewed for this research have pointed to potential advantages and disadvantages of both narrower and wider scopes (see Table 2). Generally, the merits of narrower scopes tend to overlap with objectives of system-supporting rationales while advantages of wider scopes rather reflect concerns of system-disrupting rationales.

Table 2. Potential (dis)advantages of wider and narrower scopes for CAs.

	Advantages	Disadvantages
Narrower	More actionable and considered proposals Higher policy impact	Less issues to be covered Restricting participants illegitimately Less consideration of systemic change Silo thinking
Wider	More issues to be covered More open to citizen relevancies Accommodating new and unconventional ideas	Less depth and lower trustworthiness Less relevant or useful for policymakers More cherry-picking

By definition, narrower scopes allow participants more time to go into depth when developing proposals, understanding consequences, and providing defensible justifications. For wider scopes, spectators may (rightfully) question whether all participants had sufficient time to do so. Elstub et al. (2021) found that the policy impact of the CA UK was impeded due to its large scope. Public officials doubted the representativeness of recommendations because the CA UK divided members into subgroups to cover more ground. This resulted in very small numbers of members working on each recommendation with little exchange across subgroups. Additionally, the authors argued that a large number of recommendations undermines the policy impact because policymakers are less likely to be held accountable when disregarding one controversial proposal out of one-hundred compared to one out of ten. In other words, a smaller number of proposals may reduce the likelihood of cherry-picking¹⁰ (Font et al., 2018). Interviewees think that narrower scopes are more likely to lead to actionable proposals with policy impact because they go along with rationales of system-supporting deliberation and closer coupling of CAs to policy processes. Assemblies with wider scopes run the risk of producing proposals that are less relevant or useful for policymakers.

While potentially less relevant for policymakers, wider scopes may allow participants more freedom to discuss issues they are most concerned about and develop new or less conventional policy ideas that challenge entrenched attitudes and practices, and dominant agendas. Narrower scopes risk closing down discussions and restricting participants illegitimately. They may be more likely to accommodate thinking in (sectoral) silos and less likely to foster the consideration of structural or systemic changes. Finally, narrow scopes may be seen as inefficient considering the substantial amount of resources invested in CAs.

While some of the disadvantages mentioned above are an inherent trade-off between breadth and depth other risks can be abated by taking further action. For example, organizers can counteract silo thinking by making deliberate choices on framing and information provision.

4.2. Authority

The target authority has sufficient power to act on proposals of the selected agenda.

When concerned with system-supporting objectives like policy impact, interviewees suggested to limit the agenda to only those issues where the target authority (e.g. the federal government) has jurisdictional power to work toward the adoption and implementation of proposals. That is because authorities tend to discard proposals outside of their jurisdiction as can be observed in government responses to CAs. On many issues, however, decision-making and implementation powers are dispersed across jurisdictions, e.g. levels, and ministries. In such circumstances, experts advise to draw responsible actors into the CA process early on to secure their buy-in and make later adoption and implementation more likely. Moreover, past CAs, like the French, have explicitly asked authorities to lobby for change at other political levels such as the European Union¹¹, to which the French government responded by sketching the state of negotiations.¹²

Other interviewees advised against limiting the agenda to issues where target authorities possess jurisdictional power. They argued that environmental problems often transgress political boundaries which prevents conventional political structures from dealing with such problems effectively. CAs, in this view, may counteract this weakness of current political structures. Interviewees were also concerned that limiting the agenda to target authorities’ jurisdictions will close down deliberation on more structural and systemic changes. Again, these perspectives rather overlap with goals of system-disruption.

4.3. Societal relevance

The issue is important for citizens.

Societal relevance can be defined subjectively or objectively (inter-subjectively). In subjective terms societal relevance refers to issues that citizens perceive as important to them. Other issues are important for societies in objective terms, for example because they (will) affect the well-being of a large number of people (even though they are not at the top of people’s minds). Presumably, such issues are likely to become important to citizens once made aware of the problem’s magnitude and how they are affected by it.

Some interviewees said that choosing issues of societal relevance facilitates the recruitment of participants, increases their information seeking and engagement in deliberations which ultimately bolsters the internal quality of CA processes. Others, however, have reported about CA cases of high internal quality even though the issues discussed appeared rather boring or disconnected from people’s daily lives at first sight. They stressed the importance of framing to emphasize an issue’s relevance. A risk is that too few citizens from the random sample will volunteer to participate if the topic is not relevant for them undermining the representativeness of the assembly. Moreover, when concerned about systemic impacts, interviewees highlighted that issues of societal relevance are more likely to receive higher levels of public attention.

That an issue for a CA is of high perceived importance to citizens seems to be more important if either empowering citizens and increasing the systems responsiveness to citizens’ relevancies are major objectives, or if reaching high levels of public attention and influencing public discourse are among the main goals. If, however, direct policy impact through bureaucratic and political pathways secluded from salient public discourse is the main objective, relevance to policymakers appears more important than relevance to citizens. What policymakers and citizens find important may or may not align and is highly dependent on context.

4.4. Political relevance

Policymakers see a need for change on the issue.

Interviewees strongly agreed that the policy impact of CAs likely increases when policymakers see a general need for action on the issue. CAs are argued to be particularly useful where conventional decision-making structures have not been able to address a problem effectively. Examples include issues that are in political deadlock, issues that are ‘too hot to handle’ (Willis, 2020) because politicians anticipate risks of backlash or may lose face, or issues where politicians have conflicts of interest.

From a system-supporting perspective on CAs, considering political relevance is highly important when setting the agenda. When prioritizing system-disrupting objectives, less emphasis is put on political relevance. From this perspective, CAs are welcomed when they elevate marginalized agendas instead of reinforcing discourses and solutions of those that already hold powerful positions.

4.5. Receptiveness

Policymakers welcome citizen input on the selected agenda.

Receptiveness refers to policymakers’ more general attitudes regarding citizen participation and their willingness to take CA proposals seriously on the selected agenda issue. If policymakers are not supportive of the idea of citizen input, the political impact of an assembly is likely to be low. Receptiveness can differ between issues as the responsible policymakers holding the attitudes may differ and the nature of the issue – e.g. its salience – may affect receptiveness. Ensuring both receptiveness and relevance of an agenda poses a challenge as policymakers tend to be less willing toward citizen input on highly salient issues if they perceive a risk of unfavourable outcomes – e.g. proposals that challenge their agendas or past actions. David Farrell has criticized the Irish government of assigning citizens’ assemblies to ‘rather daft issues’.¹³ This points to the question of who ought to have the right to initiate CAs (see discussion).

If receptiveness is low, e.g. because a CA was not initiated by those in power, policy impact may be higher if rules are in place on how decision makers must respond to proposals or if the assembly sparks heightened media attention and public pressure, which has been the exception. But even if powerholders commission a CA, receptiveness may still pose a challenge. It is worthwhile considering the actor constellation in the empowered political space as governments, administrations and parliaments are not unitary. On a given issue, some powerful actors may be highly receptive to CA input while others are not. Under such circumstances, strategic actions might be required to increase support among skeptics or put pressure on them, if the objective is that the CA has policy impact. Some interviewees note that policymakers and public officials unfamiliar with CAs often become less suspicious and more favourable of the process if they have opportunities to directly experience the assembly and meet participants.

4.6. Timing

There is an opportunity to affect change.

Interviewees agreed that commissioners should consider how the timing of a CA may affect its impacts (see also Zahariadis, 2007). One should avoid running a CA when decisions on the issue have already been taken. Foreseeable windows of opportunity include recent or upcoming policy processes, elections, or related events with news value like political summits or heat waves.

When aiming for immediate policy impact, it seems particularly important to align a CA with policy processes and timelines of policymakers in charge of the issue. From a system-disrupting perspective this is less important because CAs are rather seen as tools to open a window of opportunity or spark discussion on marginalized issues for which no ongoing policy processes exist.

Some noted that CAs should not be held shortly before an election because a new government that does not support the CA (results) has an opportunity to ignore its proposals.¹⁴ Contrarily, a civil society organization in Germany commissioned a climate assembly with the aim to affect elections and policy in the coalition agreement between newly elected parties. While the impact of this CA was limited¹⁵, the beginning of a legislative period is often a window of opportunity for policy change. Whether it can also be a window of opportunity for CAs remains to be investigated and is likely moderated by many other factors.

4.7. Dilemmas

Clear trade-offs must be made.

Many interviewees strongly emphasized that CAs should be focused on those aspects of issues where trade-offs between values and interests arise. Most past assemblies have led to long wish lists of proposals that do not explain how measures are to be prioritized or financed. This provides little guidance to policymakers as it does not indicate whether citizens are willing to accept higher taxes, debts, or budget cuts elsewhere.¹⁶ A citizens’ assembly on diets in Germany was recently asked to limit their number of proposals to nine, explicitly stating the aim to force citizens to discuss their priorities.¹⁷

Some said that the issue given to a CA should not be too technical because citizens may get overwhelmed and are not competent enough. Others strongly disagreed with this statement saying that no political issue is too technical for a CA. They pointed to successful cases of deliberative minipublics on genetic engineering as examples. Further clarification revealed that below the surface of these contradictory statements lies surprisingly little disagreement. Whether or not an issue is too technical depends on how it is framed, i.e. what the specific aspects of the issue are that citizens are asked to give advice on. It is less useful to ask citizens for advice on questions of technical optimization regarding a value (e.g. What type of electricity mix yields the lowest consumer prices?). A diverse group of experts can legitimately provide better answers, here. Citizen advice is useful when asked about which values matter to them (e.g. prices, distributive fairness, emissions, etc.) and how to balance these values on a particular policy issue (also see Renn et al., 2007). To sum up, CAs can deal with complex technical information but should be focused on producing actionable proposals on issues where societies face difficult dilemmas on how to act.

4.8. Legitimacy

The agenda is seen as legitimate by most groups.

A concern among interviewees and deliberative theorists is that CAs will be manipulated to serve the interests of the sponsors, especially those in political power (also see Bua, 2012). Such practices would likely undermine trust in CAs and potentially its sponsors. Therefore, many interviewees support the idea that a wide range of different social groups and interests accept (or even support) the CA and its agenda from the outset. However, interviews also showed that a fear of conflict and bad publicity can lead sponsors to choose agendas that are less relevant. Holding CAs on issues of low political and societal relevance undermines system-supporting as well as system-disrupting objectives.

Satisfying all interests when setting the agenda for a CA is likely an exception. Practitioners have used at least two strategies to address this difficulty. The first is to be transparent in the choices made, and to provide defensible justifications. Beyond that, processes like the Scottish Climate Assembly¹⁸ brought different political and social groups together to deliberate on the agenda.

From a system-disrupting perspective, acceptance by most groups may or may not be desirable depending on context. A civil society sponsor may want to demonstrate widespread support to gain recognition in empowered political spaces (as was the case for the German CA). On the other hand, conflict may increase the news value of the assembly and its proposals thereby increasing its wider impacts.

4.9. Openness

Citizens are not unjustly constrained by the agenda.

Interviewees stressed that citizens should never lose their sense of ownership of the process and their recommendations. Citizens may lose their ownership if they feel that the agenda is too constrained or if they feel that the agenda is leading them toward recommendations desired by the sponsor. This would not only be normatively illegitimate but can also undermine the perceived legitimacy of the process and thereby reduce a CA’s policy impact, or even lead to adverse effects for the sponsor.

Still, sponsors with system-supporting objectives can have defensible reasons for constraining the agendas to issues that are most politically relevant. Being transparent and allowing space for citizens to scrutinize and add to the agenda not only reduces the risk of citizens losing their sense of ownership but also increases the normative legitimacy of the process. From a system-disrupting perspective, one would prefer to give more freedom to citizens in setting the agenda. The Ostbelgien-Model, for instance, allows citizens to set their own agenda (see Niessen & Reuchamps, 2019). Experience with CAs in Denmark suggests that citizens should be supported with expert information when given the opportunity to choose their agenda (Brancaforte & Pfeffer, 2022). However, citizens in assemblies are not a unitary mass. They may experience conflict over the choice of agenda items. Then, the issue is how to decide. Majority vote? Hope of consensus? Include whatever is suggested? The German assembly on nutrition includes minority statements in their final reports to highlight issues of disagreement.¹⁹

4.10. Resource efficiency

Societal benefits of citizen deliberation on the selected agenda outweigh invested resources.

Organizing a CA requires investing a considerable amount of money and time, including that of the participants, as they often run for more than 6 weeks and engage more than 100 citizens. Therefore, some interviewees noted that citizens’ assemblies should only focus on big agenda issues, that is on issues where the assembly is likely to deliver societal benefits large enough to outweigh its costs. When planning to organize a citizens’ assembly, sponsors are well-advised to ask themselves whether less resource intensive alternatives to CAs exist (e.g. citizens’ juries) that would deliver similar benefits (Boswell et al., 2023).

It is, of course, difficult to identify (or even quantify) all the societal benefits a CA can deliver on a given issue. At least heuristically, practitioners can ask how big the problem is – e.g. how many people are affected, how many material and immaterial costs is it causing and projected to cause? – and what degree of problem abatement can be expected due to the CA given the way the agenda is framed and constrained – e.g. what kind of proposals might result, how effective would such proposals be, how likely are they to be implemented or to cause disruption via public pressure?

5. Conclusion and discussion

While setting the ‘right’ agenda does certainly not guarantee a success, a CA is unlikely to be an effective tool for democratic deliberation and decision-making with a poorly set agenda. Results underscore that design choices for CAs require making trade-offs between objectives connected to different rationales. Hammond’s (2020) distinction in system-supportive and system-disruptive rationales has proven useful to cluster different theories of change related to CAs. Those assuming that government is generally willing to act on climate change but requires support from CAs tend to see benefits in designing narrower agendas fitting ongoing policy processes and political demands, which can come at the expense of constraining deliberations and only lead to incremental policy changes. Those assuming government is unwilling to act or unwilling to instigate the levels of systemic transformations needed to meet climate targets tend to favour agendas that are less constrained by political imperatives, allowing more space to challenge existing institutions, actors, and paradigms. This, however, risks that the immediate impact on policymaking might be low (Figure 1).

Figure 1. Setting the agenda for climate assemblies: facing complex realities.

Note that most interviewees in my sample tended toward system-supporting rationales, although many considered both perspectives. The extent to which this is representative of the expert population is unknown. A higher share of activists in the sample might have led to different framings and evaluations of trade-offs. Moreover, the paper presents expert beliefs about causal relationships between CA designs and effects. This allows for theory development and deriving hypotheses, but not for testing their validity and effect sizes.

Beyond CAs, the insights of this paper may inform debates on minipublic purposes and agenda designs more generally. System-supportive and -disruptive rationales and trade-offs are not limited to climate practitioner contexts but appear in wider scholarly works on deliberative practices (Böker, 2017; Caluwaerts & Reuchamps, 2016; Courant, 2022; Curato et al., 2021; Curato & Böker, 2016; Harris, 2019; Jacquet & van der Does, 2020; Ravazzi, 2023). Paper results support arguments by Bussu and Fleuß (2023) that simple distinctions of top-down and bottom-up citizens’ assemblies are not useful because real cases are mostly hybrid. Following them, I analyzed single elements of assemblies to assess whether they facilitate rather top-down (system-supportive) or bottom-up (system-disruptive) objectives. Such fine-grained analyses can ‘be used as a “toolbox” that allows for flexible design and combination in democratic processes’ (Bussu & Fleuß, 2023, pp. 150–151), allowing practitioners to make better decisions. Thus, future studies should analyze how different elements of minipublics interact. One could compare cases where political sponsors retain full agenda control (CA Finland) to cases where citizens have full control (East-Belgium) and hybrid cases (CAs Denmark, Scotland) (see last paragraph).

For practitioners having to make agenda choices, the elephant remains in the room: Amid all the trade-offs, what are the ‘right’ decisions? I suggest the answers are not purely subjective. Rather, practitioners should reflect their assumptions and confront them with their political context. As Bussu et al. (2022) have suggested with the concept of ‘embeddedness’, a normatively productive role for a participatory process depends on the elements that a ‘good’ democratic (sub-)system is lacking (also Bächtiger & Goldberg, 2020). Is the system characterized by political deadlock that leads to aggravations of untreated problems (Mansbridge, 2012), then a productive CA role may be an instrumental one with a tight agenda aimed at breaking this deadlock. Does the system lack responsiveness to public concerns or is there a general unwillingness for climate action in government? A more disruptive agenda providing space to challenge those in power may be more desirable in this case. Importantly, climate political contexts are more complex and diverse than sketched here.

So far, I have neglected to ask how realistic it is that CAs can fulfil system-disruptive functions, at all. Without further institutionalization, disruptive minipublics may only be an exception (Courant, 2022). After all, minipublics will only be initiated if those with the power to initiate them expect to benefit sufficiently. Governments are unlikely to set agendas that challenge themselves. To embed CAs more productively within climate governance, one could introduce initiation rights for citizens (e.g. if a threshold of signatories is surpassed), or ‘automatic’ initiation mechanisms (Setälä, 2017) (e.g. if governments fail to meet their climate targets). Still, governments may simply ignore such unwanted CAs, unless they generate high levels of public attention and pressure – which has been the exception for past assemblies. Therefore, new initiation mechanisms would have to be paired with other provisions that either increase the accountability of those in power toward the assembly (e.g. through certain requirements to respond), or with processes shifting decision-making power to citizens like follow-up referendums, as practiced in Ireland (Farrell et al., 2023) or Oregon (Gastil & Knobloch, 2020).

Finally, more research is needed on the procedural dimensions of CA agenda-setting, such as whom to involve in agenda decisions and with how much power (for a start, see Pfeffer, forthcoming). The governance structures of past CAs have varied considerably (Carrick, 2022). In Scotland, a stewarding group comprised of stakeholder representatives, members of parliament, civil servants, participation experts, and climate scientists set the remit.²⁰ In Denmark, CA members actively contributed to shaping the agenda, and members of the German assembly on nutrition chose the sub-themes for deliberation themselves.²¹ And in Finland, the CA was constrained to assess several proposals formulated by government.²² These differences matter as actors vary in their interests and knowledge. For example, policymakers and civil servants know what is politically relevant, timely, and within their authorities – knowledge that CA members typically lack. However, politicians are also driven by concerns about re-election and career progression – restraints that CA members do not need to worry about. Future research should investigate the possibilities of designing embedded CA processes that both allow for productive disruptions while being sufficiently tailored to policy processes to have immediate impacts on policymaking (see Courant, 2022 for a conceptual start).

Ethics approval

In accordance with the procedural instructions of Ethical Review at the authors' home institution, a formal review was deemed unnecessary because all research participants gave informed consent in both oral and written form, topics discussed were non-sensitive, and the highly educated participants were familiar with expert interviews.

Acknowledgements

I would like to give special acknowledgement to Amélie Schenavsky and Balthasar Klingenhage for their contributions to early versions of the interview guide. Moreover, I thank Graham Smith, Kristinn Már Ársælsson, Ortwin Renn, panel participants at the ECPR 2022, and peer-reviewers for their thoughtful comments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability

The data that support the findings of this study are available on request from the author. The data are not publicly available due to their containing information that could compromise the privacy of research participants.

Additional information

Funding

This work was supported by the Knowledge Network on Climate Assemblies through funds granted to a knowledge development project; and the Heinrich-Böll-Stiftung through a doctoral scholarship.

Notes

1 This article is informed by an earlier and less detailed version by the author aimed at a non-scientific audience, see Pfeffer (2022).

2 See https://knoca.eu/national-climate-assemblies/.

3 See https://www.lemonde.fr/en/les-decodeurs/article/2023/05/26/france-s-short-haul-domestic-flight-ban-a-measure-lacking-substance_6028097_8.html.

4 See footnote 2.

5 Next to climate-related rationales, there are many democratic and political objectives. Scholars have produced a rich literature proposing criteria for assessing whether minipublics like CAs ‘improve’ democracies, Warren (2017), including whether they generate factually informed preferences, Mansbridge et al. (2012), enhance critical scrutiny in public discourse, Curato and Böker (2016), lead to tangible political or societal consequences, Dryzek (2010). Educational objectives, endorsed by many politicians, aim to (re-)connect elites and citizens by sensitizing citizens about the complexities of political work, increasing knowledge, and promoting civic values. Political actors often have strategic objectives, such as gaining voter popularity or suppressing opposition. Some are critical of CAs, either based on well-reasoned arguments (e.g. lack of electoral legitimization and accountability mechanisms), Lafont (2019), or simply because they are potentially threatening their power.

6 See https://knoca.eu/france-citizens-convention-on-the-climate/.

7 As many experts have taken on multiple roles and worked on assemblies in multiple countries, it is difficult to provide precise characterizations.

8 See https://knoca.eu/about-us/.

9 See footnote 2.

10 Cherry-picking means that sponsors only adopt those proposals that fit their pre-existing agendas.

11 See the French CA, for example: https://propositions.conventioncitoyennepourleclimat.fr/objectif/mieux-prendre-en-compte-les-emissions-de-gaz-a-effet-de-serre-liees-aux-importations-dans-les-pollutions-europeennes/.

12 https://www.ecologie.gouv.fr/suivi-convention-citoyenne-climat/?id_rubrique=4#affiner, PT 9.1.

13 https://www.irishtimes.com/opinion/we-may-have-overdone-it-on-citizens-assemblies-1.4803375.

14 Unexpected early elections have interfered CA impacts in Spain and the city of Berlin.

15 It is unlikely the process impacted the elections due to low media attention. Regarding policy impact, there are many overlaps between the CA results and the coalition agreement but expert interviews and informal conversations with involved policymakers revealed that these overlaps were not caused by the CA. The coalition agreement rather reflected a pre-existing policy consensus between the coalition parties independent of the CA.

16 Processes like participatory budgeting address such trade-offs by design. This may be one of many reasons why participatory budgeting proposals achieve higher degrees of implementation compared to other processes, see Font et al. (2018).

17 See Mehr Demokratie, nexus, IPG, and ifok. (2023, p. 7). Deliberative Committees in Belgium have also begun to limit the number of recommendations, see https://www.bertelsmann-stiftung.de/en/our-projects/democracy-and-participation-in-europe/shortcut-archive-1/shortcut-9-deliberative-committees-a-new-approach-to-deliberation-between-citizens-and-politicians-in-brussels.

18 See https://www.gov.scot/publications/scotlands-climate-assembly-operations-administration-arrangements/pages/10/.

19 See Mehr Demokratie et al. (2023).

20 See footnote 17.

21 See https://knoca.eu/denmarks-climate-assembly/, and footnote 18.

22 See https://knoca.eu/finlands-citizens-jury-on-climate-actions/.