2. Models of representative deliberative processes
Numerous models of representative deliberative processes have been developed, tested, and implemented across the world. They can be clustered into four types of purpose:
(1) informed citizen recommendations on policy questions;
(2) citizen opinion on policy questions;
(3) informed citizen evaluation of ballot measures, and
(4) permanent representative deliberative models.
This chapter’s first section introduces 12 models of representative deliberative processes, broken down by the types of purpose. The models described are: Citizens' Assembly; Citizens' Jury/Panel; Consensus Conference; Planning Cell; G1000; Citizens' Council; Citizens' Dialogue; Deliberative Poll/Survey; World Wide Views; Citizens' Initiative Review; the Ostbelgien Model; and the City Observatory.
The second part of this chapter outlines how to choose between different models depending on the purpose, complexity, issue and other factors. The chapter concludes with a discussion of combining features of different models.
Introduction
Drawing on the new empirical research collected for this report and broader theoretical research, this chapter introduces 12 models of representative deliberative processes (referred to interchangeably as deliberative processes for shorthand), grouped by four types of purpose. It defines their design characteristics and outlines how to choose between the different models depending on the purpose, complexity, issue and other factors.
Over the years, due to the combined efforts of policy makers, academics, and civil society, numerous models of representative deliberative processes have been developed, tested, and implemented across the world. As their use has spread, some models have come to be named differently depending on the country, but remain essentially similar. For instance, Reference Panels in Canada and Citizens’ Juries in Australia fall under the same model type, despite their differences in nomenclature. These divergences, as well as others, are partly down to political culture and history, which are discussed in the relevant sections.
Overall, the choice of deliberative models has so far depended on the familiarity with the model and experience using it, leading to preferences in different countries for specific models. However, their widespread use signals their universality and potential applicability in different national and local contexts.
The deliberative models presented in this chapter are not necessarily exhaustive. Each model shares the essential phases of quality representative deliberative processes: learning, deliberation, and the development of collective recommendations.
Empirical examples of representative deliberative processes in this report meet the three criteria of: (1) being commissioned by public authorities; (2) participants being randomly selected and demographically stratified; and (3) one day or longer of face-to-face deliberation.
Overview of different models
The models can be characterised by four types of purpose:
1. Informed citizen recommendations on policy questions: These processes require more time (on average a minimum of four days, and often longer) to allow citizens adequate time and resources to develop considered and detailed collective recommendations. They are particularly useful for complex policy problems that involve many trade-offs, or where there is entrenched political deadlock on an issue.
2. Citizen opinion on policy questions: These processes require less time than those in the first category, though still respect the principles of representativeness and deliberation, to provide decision makers with more considered citizen opinions on a policy issue. Due to the time constraints, their results are less detailed than those of the processes designed for informed citizen recommendations.
3. Informed citizen evaluation of ballot measures: This process allows for a representative group of citizens to identify the pro and con arguments for both sides of a ballot issue to be distributed to voters ahead of the vote.
4. Permanent representative deliberative bodies: These new institutional arrangements allow for representative citizen deliberation to inform public decision making on an ongoing basis.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Models for developing informed citizen recommendations on policy questions
Citizens’ Assembly
The Citizens’ Assembly is considered as the most robust and elaborate model of representative deliberative processes (Escobar and Elstub, 2017). It was first introduced in Canada in the early 2000s – 2004 in British Columbia and 2006 in Ontario – to address the question of electoral reform. The first Citizens’ Assembly was organised in response to the need to create a platform where ordinary citizens, rather than political elites (who may have been influenced by party loyalties), could contribute to the design of a new electoral system for British Columbia.
Citizens’ Assemblies, as characterised in this chapter, have been mostly used to address questions to do with institutional setup and constitutional changes. They have also tended to be used in contexts of political tension. For example, the 2016-2018 Irish Citizens’ Assembly was set up to solve a political and social divide on contentious issues unresolved in years – same-sex marriage and abortion. The 2019-2020 French Citizens’ Convention on Climate has been an answer to social mobilisation – a direct outcome of the Yellow Vest movement and a sequel to its first response, the Great National Debate.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Given the reasons for their use, Citizens’ Assemblies tend to involve more participants and last longer than the other representative deliberative models. On average, Citizens’ Assemblies gather 90 citizens (ranging from 36 to 161, although four of the six examples have 99 participants or more) for 18.8 days. Typically, these are spread over numerous weekends. In the cases collected by the OECD, the average duration is around 11 months.
They begin with a learning phase, where citizens become familiar with the policy question and consider a range of perspectives presented by experts, stakeholders and affected groups, a diverse mix of whom present to the participants in person and answer their questions. It is also common for citizens to be able to request additional information, experts or stakeholders if they feel they are missing information or need additional clarifications. This is one of the reasons why meetings tend to be spread out over numerous weeks, to allow the organisers to invite additional experts and prepare extra information if needed (see Chapter 4 for more details).
Often an independent advisory group of researchers is formed prior to the assembly with a mandate to prepare a diverse information and evidence base for this purpose. Assembly members also consider inputs from other citizens, either by holding citizen hearings and calling for online submissions, or drawing from various citizen consultation and engagement processes that have been done in preparation for the assembly.
Learning and consultation is followed by citizen deliberation, when evidence is discussed, options and trade-offs are assessed, and recommendations are collectively developed. The process is carefully designed to maximise opportunities for every participant to express their opinion and is led by impartial trained facilitators.
The final set of recommendations is voted on by all participants, most commonly by a majority vote, resulting in a detailed report and often a minority report, which acknowledges other opinions that were expressed but did not achieve majority consensus. Final recommendations are made publicly available and presented to the government authority. The government authority responds to recommendations providing feedback to the participants and the broader public.
Lengthy and elaborate, Citizens’ Assemblies typically attract more media attention than other types of processes. Since all the learning material and information that Citizens’ Assembly members access is made public, this representative deliberative model creates an opportunity for enriching the wider information debate to encourage widespread and informed deliberation on a key policy question well beyond the group of selected Citizens’ Assembly members (Suiter, 2018; Fournier et al., 2011; Warren and Pearse, 2008).
The Irish Citizens’ Assembly involved 100 randomly selected citizen members who considered five important legal & policy issues: the 8th amendment of the constitution on abortion; ageing populations; referendum processes; fixed-term parliaments & climate change. The Assembly’s recommendations were submitted to parliament for further debate. Based on its recommendations, the government called a referendum on amending the 8th amendment and declared a climate emergency.
More information can be found here: https://www.citizensassembly.ie/en/.
Note: This document and any map included herein are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Citizens’ Jury/Panel
Used at all levels of government, Citizens’ Juries and Panels have been initiated to address a broad range of policy questions, the most common ones being infrastructure, health, urban planning, environment, and public services. Most of them have been ad hoc, but there is also one institutionalised model of an ongoing Panel. It has two examples – the Toronto Planning Review Panel (Box 2.2) and the Greater Toronto Hamilton Area Metrolinx Transport Panel (see Chapter 6 on institutionalisation for more details about both).
Citizens’ Juries and Panels follow the same learning, deliberation, and decision-making phases as Citizens’ Assemblies, but more concisely. They are, to date, the most adapted of representative deliberative models, and three main sub-categories have emerged over time:
1. processes that have taken place over consecutive days;
2. processes where meeting days are spread out over numerous weeks, and
3. ongoing panels over much longer periods of time (e.g. two years).
Citizens’ Juries and Panels have often been combined with a rich array of other citizen participation practices that precede the jury or are conducted as parallel citizen engagement activities. These include community meetings, surveys, and online calls for proposals, advisory committees, community discussions, public consultations, focus groups, neighbourhood meetings, and others.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Processes that have taken place over consecutive days
The Citizens' Jury was developed in the United States by Ned Crosby and the Jefferson Center in 1971. Ned Crosby designed it while writing a doctoral dissertation on social ethics. His goal was to invent a process that would enhance reason and empathy among citizens as they discussed a public policy matter or evaluated candidates (Jefferson Center). The initial design and method follows a rigid model, and causes some confusion as many processes labelled as Citizens' Juries in other countries do not follow the same strict design criteria of the initial model.
Distinctive characteristics of the Jefferson Center Citizens' Juries are that they are usually smaller than average – between 12 to 24 people – and they typically run three to six days consecutively (Jefferson Center). While this approach was developed in the United States (US), it has been replicated in other places, including examples in Australia, Canada, Denmark, France, Korea, Spain, and the UK.
Processes where meeting days are spread out over numerous weeks
In contrast, similar processes called Reference Panels in Canada, pioneered by MASS LBP, evolved from the experience with Citizens' Assemblies in British Columbia and Ontario in the late 2000s. During this same period (and without awareness of one another at the time), the newDemocracy Foundation in Australia was separately developing a similar deliberative model to MASS LBP's, calling its processes Citizens' Juries.
The Canadian and Australian Reference Panels and Citizens' Juries tend to involve larger groups of participants (usually around 36 to 45) and the meetings are spread out over numerous weekends, based on the view that this is crucial for the learning process and for quality deliberation. They also began the trend of a new and rigorous two-stage method for random selection. MASS LBP coined the term "civic lottery" to describe it, which is now widely used. The civic lottery involves an initial step where a large number (typically 10,000-20,000) letters are sent by post to a random portion of the population. The letter contains an invitation to participate in the Reference Panel or Citizens' Jury, often signed by the public authority commissioning it, and contains key information about the purpose, remit, duration, meeting dates, and frequently asked questions. Among those who accept this invitation, a second step involves random selection with demographic stratification to ensure that the final make-up of the group reflects a wide cross-section of society. The details of this process are available in MASS LBP's handbook on How to Run a Civic Lottery (2017) and are discussed in more detail in Chapter 4.
In the UK , there was a peak in the use of processes called Citizens' Juries, which more closely resembled the Jefferson Center's approach, in the late 1990s/early 2000s. However, these processes were not particularly well-regarded in terms of their design integrity and impact, and their use more or less stopped at around the time of the 2008 financial crisis (Chwalisz, 2017). Policy makers ignored the method for close to a decade. When the deliberative wave picked up again in the UK in the late 2010s, the previous use of the term Citizens' Jury carried negative connotations. It is one of the likely reasons why the term Citizens' Assembly has been used to describe many of the most recent processes that are in fact more similar to Citizens' Juries and Reference Panels as practiced elsewhere, rather than to the historical precedent of the characteristics of a Citizens' Assembly as described in this chapter.
Some of the UK examples have been closer to the Jefferson Center’s model, while others are more similar to the MASS LBP/newDemocracy Foundation approach depending on the practitioner. However, the civic lottery was not used in the UK before 2019. In 2019, the UK’s Innovation in Democracy programme was launched, through which the Citizens’ Jury model has been implemented in several local level deliberative processes across the country.
Moreover, in Poland, the Citizens' Panels ("panel obywatelski") that have taken place are closely aligned to with the practices of MASS LBP in Canada and the newDemocracy Foundation in Australia, although they tend to be slightly larger in size (around 60 participants). Participants are chosen through a civic lottery and the meetings are also spread out over numerous weeks. In English, the Polish processes are often cited or referenced as Citizens' Assemblies, however, in this study they are categorised under the Citizens' Jury/Panel model due to their design similarities.
Other terms that have been used to describe processes that meet the characteristics of a Citizens' Jury/Panel are Community Panel and People's Panel. In non-English-speaking countries, there are other variations.
Ongoing panels over much longer periods of time (e.g. two years)
Finally, the third sub-category of Citizens’ Juries/Panels refers to an ongoing representative deliberative body for a longer period and on multiple issues related to one policy area. As of early 2020, it has been used only in Canada and run by MASS LBP, with many of the same characteristics of a Reference Panel in terms of average number of participants (around 30), selection through a civic lottery, an in-depth learning phase, deliberation moderated by skilled facilitators, and ultimately the provision of informed inputs to policy makers.
The example of such an ongoing deliberative body in this study is the Toronto Planning Review Panel (TPRP) 2015-2017. A second iteration of the panel also took place from 2017-2019, with a new group of randomly selected Toronto residents. The remit of the TPRP is to provide informed inputs on a regular basis on planning issues to the City’s Chief Planner and Planning Division. At the time of writing in early 2020, a panel with similar characteristics is operating on transportation issues in the Greater Toronto and Hamilton Area, commissioned by Metrolinx, the regional public transport authority.
It is understandable that there has thus been much debate and confusion about terminology among practitioners and academics, as the same terms have been applied to different processes, largely driven by different political contexts. These differences are acknowledged and the OECD has attempted to group the processes with similar design characteristics, regardless of what they are called, to allow for a more accurate comparative analysis. For this reason, five processes1 that were called "Citizens' Assemblies" (three in the UK and two in Canada) have been reclassified as Citizens'/Juries Panels for the analysis of deliberative models in this report (see Annex B for full methodology details).
Citizens’ Jury/Panel that has taken place over consecutive days
Forest of Dean District Citizens Jury (2018) took place in the United Kingdom. The National Health Service bodies commissioned a Citizens Jury that provided residents the chance to evaluate prospective hospital locations and choose the one that best suits citizens’ needs.
More information can be found here: https://jefferson-center.org/forest-of-dean-citizens-jury/
Citizens’ Jury/Panel where meeting days are spread out over numerous weeks
Melbourne People's Panel (2014) in Australia provided 43 randomly selected citizens with an opportunity to contribute to the 10 Year Financial Plan of the City of Melbourne and provide their recommendations on the allocation of resources. This was the largest city budget opened up to a deliberative process, reaching $400 million Australian dollars.
More information can be found here: https://www.newdemocracy.com.au/2014/08/05/city-of-melbourne-people-s-panel/
Ongoing Citizens’ Jury/Panel
Toronto Planning Review Panel (2015-2017)
The Toronto Planning Review Panel was an ongoing deliberative body, embedded into the city’s planning division, which enabled ongoing citizen input on the issues of planning and transportation. Its members served two-year terms, after which time a new cohort was randomly selected to be representative of the Greater Toronto Area. A group of 28 randomly selected residents from all parts of the greater Toronto area met for 11 full-day meetings from 2015-2017. Prior to deliberation, participants met for four days of learning and training. A similar panel was appointed for the 2017-2019, this time consisting of 32 randomly selected citizens.
More information can be found here: http://bit.ly/3brvnxv.
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Consensus Conference
The Consensus Conference was developed in Denmark in 1987 by the Danish Board of Technology. It is based on a model of technology assessment that originated in the USA during the 1960s, which did not include elements of citizen deliberation at the time, as citizens were solely responsible for choosing the expert panel that would deliberate amongst themselves.
The model of Consensus Conferences featured in this report tends to bring together, on average,16 randomly selected citizens for four days, usually Friday to Monday. The learning stage happens during a preparatory weekend. Participants dive deeper into the policy question and identify a range of questions they would like to ask the expert panel – comprised of scientists, practitioners, and policy makers. During the first day of the conference, the expert panel presents their perspectives, and citizens question the panel’s positions (The Danish Board of Technology, 2006).
This is followed by citizen deliberation and writing recommendations during the next half day. Consensus Conferences are specific as citizens usually have to reach a consensus on the recommendations they have produced, indicating the points for which 100% consensus was reached. During the final day, citizens present their recommendations to the panel of experts and politicians.
This design enables an immediate response from the experts to citizens, as citizen suggestions are then discussed. In addition, press and broader members of society are invited to participate in the final phase, making citizen recommendations public and opening them up for wider deliberation and debate.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
To date, the model has been used over 50 times all over the world to address policy questions related to new technologies, health, and environment (The Loka Institute, 2013). Only the Consensus Conferences that have met all three criteria for inclusion (impact; representativeness, and deliberation) have been included in this study.
Fourteen randomly selected citizens were brought together to provide feedback and reach a consensus on the use of gene technology in food production. Citizens recommended an establishment of establishing the Gene Technology Office, that is in charge of ensuring the labelling of genetically modified foods. They further recommended a licence fee for companies selling genetically modified foods.
More information can be found at: http://www.abc.net.au/science/slab/consconf/dinner.htm
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Planning Cell
The Planning Cell is a representative deliberative process model developed by Peter Christian Dienel at the University of Wuppertal in 1970s in Germany, where it has been used ever since at both local and regional levels to address various policy questions, most often regarding urban planning issues. Having observed a one-sided relationship between government and citizens, Peter C. Dienel designed the Planning Cell as a tool for citizens to express their opinions and participate in decision making. What makes this model particular is that it usually consists of numerous planning cells happening running in parallel to one another, sometimes considering different elements of the same overarching problem.
The Planning Cell has also been adapted in Japan, where it has been mostly used on at the municipal level and called Citizen Deliberation Meetings – Shimin Tougikai (Nagano, 2020). The Japanese model is particular, however, as it usually has only one Planning Cell happening running at a time.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
The planning cell typically gathers 24 randomly selected citizens for 3.2 consecutive days to develop a citizens’ report – a collective position on a policy question discussed. It also follows the core phases of a representative deliberative process (learning, deliberation, decision making). It has a clearly defined structure of the learning stage, with 1.5 hour-long information sessions dedicated to different sub-topics relevant to the main policy question (Nexus, 2019).
In addition, the Planning Cell has some flexibility when it comes to citizen deliberation, as participants deliberate in small groups without facilitators. This is a key distinction to Citizens' Assemblies and Citizens' Juries/Panels, which involve skilled facilitators. Not having skilled facilitators makes a Planning Cell less costly to organise, but it also presents a challenge to the quality of deliberation, as it is not assured that more confident individuals will not dominate discussions.
While the average duration of Planning Cells in this study is 3.2 days, with many lasting at least four days, there are numerous examples of shorter Planning Cells that last only two days. These instances tend to be for less complex policy issues, and thus their results after a shorter period of learning and deliberation would fall closer to the second category of purpose – citizen opinion on policy issues.
A cable car for the citizens of Wuppertal (2016)
Forty-eight randomly selected citizens were brought together to discuss the possibility of building a cable car. Citizens met for four full days and engaged in learning and deliberation. They listed arguments for and against the cable car, and concluded by recommending the local government to conduct a thorough cost-benefit analysis and funding options before making a decision.
More information can be found: https://www.wuppertal.de/microsite/buergerbeteiligung/abgeschlossene_projekte/content/seilbahn.php
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Models for capturing citizen opinion on policy questions
G1000
The G1000 is a deliberative model designed as a platform for citizens to collectively develop a broader vision for a municipality or to address a specific question (G1000, 2019). It originated in the Netherlands, where the founders were inspired by the G1000 process that took place in Belgium in 2011 (which was quite different to from this model and was developed as a bottom-up, grassroots initiative). To date, it has been used by local public authorities in the Netherlands and Spain.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
The G1000 process consists of three consecutive phases: the Citizens' Summit, the Citizens' Forum, and the Citizens' Assembly, resulting in a Citizens' Decision. Citizens make up 75% of total participants, the other 25% being equal parts civil servants and/or politicians, and employers (see Chapter 4).
During the first phase, the Citizens' Summit, 150-1,000 randomly selected citizens come together for a whole day with no agenda set beforehand: participants lead themselves. Participants engage in an open dialogue, creating a joint vision and defining various solutions for the question at hand.
In the second phase, the Forum, during multiple evenings, participants work together in small groups on the ideas of the Summit to come to concrete proposals. During this phase, participants gather expertise in various ways. The work of the groups is open to reactions and responses from all residents of the municipality.
Finally, participants reconvene in the Citizens' Assembly, again a gathering of one day, to deliberate over the final proposals presented by the working groups, and to decide which proposals are acceptable to the whole.
A majority vote is required for a proposal to be included in the Citizens’ Decision, the final result of the deliberations and vote, which is first signed by all participants and then handed over to the chair of the local or regional council or parliament (G1000, 2019). Until now, all Citizens’ Decisions have been fully accepted by local and regional councils.
Around 250 randomly selected citizens were brought together with civil servants, employers and experts for four days and discussed strategies of how can the city of Steenwijkerland become energy neutral.
More information can be found at: https://g1000.nu/project/g1000steenwijkerland/
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Citizens’ Council
The Citizens’ Council is a model of representative deliberative process that has been used most frequently in Austria on the local and regional levels to address a wide range of policy questions, mostly environmental concerns and public services.
The Citizens’ Council (or Bürgerrat in German) was developed in the Austrian state of Vorarlberg and is based the model of the Wisdom Council, created by Jim Rough (Asenbaum, 2016). As it became prominently used by local governments in Austria, the Citizens’ Council is today a more established model than its predecessor, the Wisdom Council. The model was designed to address community issues in a quick and inexpensive manner, strengthening community ties along the way.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Citizens’ Councils are typically composed of 15 randomly selected citizens and last 1.7 consecutive days on average. As a shorter process, it has less distinct learning and deliberation phases, which are usually intertwined. The first part of the process allows participants to identify issues of public interest to be discussed by the Citizens’ Council within the proposed subject, and there is no strictly predetermined remit. In practice, this is less often the case and there is a clearly defined problem to be addressed.
During the next step, citizens engage in facilitated deliberation, develop solutions to the problems identified, and produce collective recommendations (Partizipation.at, 2019). A distinguishing feature is dynamic facilitation, where the facilitator encourages participants to speak their minds without having to follow a strict agenda or process. This creates a safe place for everyone to express themselves, which can lead to openness, inclusion, and creative solutions (Center For Wise Democracy, 2019).
Recommendations are then presented and discussed with the broader public in a Citizens’ Café, open to anyone. Finally, the Citizens’ Council’s recommendations are presented to the local government and a small group of participants are assigned to follow up with the government regarding the recommendations’ implementation (Partizipation.at, 2019).
As there is no separate learning phase, Citizens’ Councils do not require many resources and long preparation, and are well-suited to be used as a helpful way to periodically give citizens an opportunity to bring salient issues to the attention of local or regional government.
The state government of Vorarlberg brought together 30 randomly selected citizens for one and a half days to develop principles and priorities in the field of mobility and transport for the state of Vorarlberg for the next ten to fifteen years. Following the Citizens’ Council, a Citizens’ Café took place, where the broader public could learn about the recommendations produced and discuss them with politicians and public administration.
More information can be found at: https://vorarlberg.mitdenken.online/buergerrat
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Citizens’ Dialogues
Citizens’ Dialogues are less intensive, often two-day deliberative processes. Their use is widespread across countries and at all levels of government. They often address more than one policy question and, on average, gather 148 randomly selected citizens, though the size can vary greatly – from 18 to 499 participants. Usually they have very brief learning and deliberation stages and include expert panel discussions.
They are more suited to inform citizens on policy issues and gather their broad ideas and reactions, rather than resulting in detailed recommendations for policy makers. Often part of a broader citizen engagement strategy, Citizens’ Dialogues are also sometimes called Citizens’ Summits, Deliberative Events, Citizens’ Forums, and Deliberative Workshops.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
The federal government of Canada brought together 35 randomly selected citizens for three days to discuss the future of Canada’s energy policy. The process was complemented by five regional dialogue events (two days each). Citizens recommended the following:
1. Developing new forms of governance and oversight for energy issues.
2. Investments in clean technology research and innovation to build the new energy economy.
3. Incentives to accelerate the adoption of existing green or low-carbon energy technologies.
4. Regulations to protect the environment or reduce greenhouse gas emissions.
5. Investing in energy infrastructure that serves Canada and its communities.
6. Addressing impacts on Canadians during changes to energy economy.
More information can be found at: https://canadaenergyfuture.ca/resources/
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Deliberative Poll/Survey
A Deliberative Poll or Survey is a process that aims to capture citizen opinion change on a policy question before and after participants have been exposed to learning and deliberation. James Fishkin developed and patented the Deliberative Poll in the United States at Stanford University in 1988. The idea of the method first appeared in an academic article and was inspired by ancient Athenian democracy and Gallup public opinion polling methods (Fishkin & Luskin, 1988). As Fishkin patented the Deliberative Poll, similar approaches that do not follow the exact methodology tend to be called Deliberative Surveys.
There have been many more Deliberative Polls/Surveys conducted than are included in this study, as the vast majority of them have not been commissioned by a public authority, and have been conducted rather as academic experiments. The model has been used to address policy questions linked to public decision making in countries such as Brazil, China, Italy, Japan, Korea, and Mongolia.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
On average, the Deliberative Polls/Surveys in this study have gathered 226 citizens (ranging from 62 to 669 participants) over 1.6 consecutive days. Participants are randomly selected and undergo the first opinion survey to measure their initial attitudes towards a policy question. After plenary sessions with citizens posing questions to experts and moderated dialogues in smaller groups, a second opinion survey is conducted to capture citizens’ opinions after they have carefully considered the policy question (Center for Deliberative Democracy, 2019). Opinion changes are analysed, made public, and presented to the government authority.
Due to its scientific approach, this model of deliberative engagement is very rigid in terms of design and does not leave any room for participants to influence the process. It is better-suited to identify citizen opinion changes rather than produce detailed recommendations or extend decision-making influence to citizens.
Box 2.8. Deliberative Poll on Construction of Shin-Gori Nuclear Reactors (2017)
A Deliberative Poll took place in Korea with regards to the construction of Shin-Gori Nuclear Reactors No. 5 & 6. The government convened 471 citizens for three days with a mandate to decide whether to resume the construction of the nuclear power plants. Citizens recommended to resume the construction, and the government implemented their recommendation.
More information can be found at: https://cdd.stanford.edu/2017/proposed-deliberation-in-south-korea-on-closing-two-nuclear-reactors/
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
World Wide Views
World Wide Views is a model of representative deliberative process on an international scale. The method was first developed by the Danish Board of Technology in the lead-up to the 2009 climate COP15 in Copenhagen, with an aim of gathering citizen opinions from across the world in a way that was easy to implement, inexpensive, and consistent in every country.
To date, it has predominantly been used to address policy issues negotiated at the United Nations Climate and Biodiversity Conferences of the Parties (COPs). The process can be commissioned by an international organisation to help inform global agreements and summits, but could also be applied on a national scale.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Each World Wide Views initiative has involved numerous national and regional Citizens’ Dialogues, each gathering around 100 participants. The number of Citizens' Dialogues has varied widely, from 34 in 2012 to 97 in 2015. Although stratified, their selection is not exclusively random and includes self-selected volunteers, depending on the country. The process takes place for one full day at the exact same time across the world and follows the same format over the course of the day. There is no cross-border deliberation; what results is an aggregate of individual country perspectives (for more details see Chapter 7).
Four to five themes relating to the core policy question are identified prior to the World Wide Views event, and informational videos are prepared for each of those themes. During the day, each theme is introduced with the exact same video material across all participating countries, followed by moderated participant deliberation in small groups (World Wide Views, 2019). Each short session is concluded by individual citizen votes on the questions regarding the issue discussed.
At the end of the day, the breakdown of votes of all the participants across countries is available on an online platform for policy makers and broader society to access (World Wide Views, 2019). This model of deliberative engagement requires precise co-ordination among international partners carrying out the process, and has a strict and clear structure.
Due to its global dimension and large number of participants, it has the potential to generate media interest and initiate broader debate on the policy question at hand. However, as a short process, it is better suited to give policy makers a snapshot of citizen opinion from a range of countries on policy options already on the negotiation table, rather than informed, detailed recommendations that take into account national and local contexts.
The United Nations Framework Convention on Climate Change (UNFCCC) Secretariat, together with partnering citizen participation organisations, initiated a World Wide Views process in preparation to the COP21 United Nations Climate Change Conference. Around 10,000 randomly and self-selected citizens were brought together in 76 national panels across different countries aiming to gather citizen views on international climate change and energy policy issues. All the panels took place on the same day and the results of citizens’ opinions were immediately publicly available publicly for dissemination to policy makers involved in the UNFCCC negotiations.
More information can be found at: http://climateandenergy.wwviews.org/
Model for informed citizen evaluation of ballot measures
Citizens’ Initiative Review
In contrast to the previously described models that aim to develop citizen recommendations for policy makers, the Citizens’ Initiative Review (CIR) is a deliberative process that was designed for a representative group of citizens to evaluate a proposed ballot measure, providing informed arguments for both sides of the issue to all voters with their ballot papers.
To date, the Citizens' Initiative Review has been implemented and institutionalised in the United States in the state of Oregon, where the model was developed in the early 2000s by Healthy Democracy (and the founder of the Citizens’ Jury method, Ned Cosby, along with his colleague, Pat Benn). Governments have also piloted it in the state of Arizona, as well as the Swiss city of Sion. Additionally, academia and civil society-led pilots have been implemented in Colorado, Massachusetts, California, and in Korsholm, Finland.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Citizens’ Initiative Reviews typically gather 22 randomly selected citizens for 4.4 consecutive days, on average. As CIRs have been altered and adjusted over the years, their duration has also changed, and the current standard is four days. The two European initiatives, in Finland and Switzerland, were both held over two consecutive weekends rather than consecutive days. Prior to the first meeting, citizens have no information on what policy question they will be addressing. Due to political pressures, organisers do not prepare briefing documents in advance. Rather, citizens receive all testimony directly from campaigns and experts during the review. The process begins with a training programme for participants, providing them with the fundamentals of deliberating and evaluating information.
The following stage is learning and evaluation. Participants assess written evidence submitted by opponents and proponents of the ballot measure, and question both campaigners and independent experts. They then add to, edit, deliberate on, and prioritise all the evidence collected. The editing and refining information phase is carried out in smaller groups where participants are invited to discuss and draft evidence statements, examine costs, benefits and trade-offs of the proposed ballot measure (Healthy Democracy, 2019). Finally, they draft a collective statement that includes the most important information for all voters to know. Participants also select the strongest evidence for and against the measure, and then explain why each piece of evidence is important to one side or the other.
Their final statement is presented publicly in the press conference to the wider public and is included in the voters’ pamphlet, which reaches every voter across the state. Compared to the other representative deliberative models, the CIR is less visible publicly until it has finished. The final result of the CIR is not addressed to the government, but rather to fellow citizens, helping them make better informed choices when it comes to voting on a ballot measure. The method can be a powerful tool to help counteract the spread of misinformation and disinformation ahead of a vote.
The Oregon state government brought together 20 randomly selected citizens for four days to conduct an in-depth study of a ballot question on corporate tax and share their findings with their fellow voters. After citizens became acquainted with the arguments for and against the proposed measure, arguments they found most important and convincing were included in a voter’s guide and delivered to every voter across the state.
More information can be found here: https://sites.psu.edu/citizensinitiativereview/files/2015/01/Assessment-of-the-2016-Oregon-CIR-zmzb9i.pdf
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Models of permanent representative deliberative bodies
Ostbelgien Model
The Ostbelgien Model is a recently developed model that combines a permanent representative deliberative body with the ongoing use of representative deliberative processes. The permanent deliberative citizens’ body is called a Citizens' Council and it has a mandate to initiate and decide the issues of ad hoc Citizens’ Panels. It originated in the German-speaking Community of Belgium, Ostbelgien, and was established through a piece of legislation, for which all political parties unanimously voted in favour, in February 2019 (Foundation for Future Generations, 2019).
The Citizens’ Council’s inaugural meeting took place on 16 September, 2019. The model was designed in 2018 by a group of practitioners, academics and experts of deliberative and participative processes, which included Claudia Chwalisz from the OECD. They were brought together and given such a mandate by the parliament of the German-speaking Community of Ostbelgien.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
In this model, 24 randomly selected citizens form the Citizens’ Council. They have a mandate to represent fellow citizens for one and a half years. The first 24 members are comprised of three different groups: six are randomly selected among the participants of a previous Citizens’ Panel that took place in the region; six are politicians – one from each political party, and twelve are randomly selected citizens from the population of Ostbelgien. Every six months, one third of the cohort is rotated out, to be replaced with randomly selected citizens. The politicians will be the first to be rotated out and will be replaced by citizens selected through a civic lottery. This is to allow for some continuity, but also to ensure that the Citizens’ Council does not become a body of people who have become professionalised and prey to some of the same problems as elected politicians.
The Citizens’ Council has the power to set its own agenda and initiate up to three ad hoc Citizens’ Panels on the most pressing policy issues of their choice. Citizen proposals that have the support of at least 100 citizens, as well as proposals of parliamentary groups or the government can also be submitted for the consideration by the Citizens’ Council (Parliament of the German-speaking Community of Belgium, 2019). Each Citizens' Panel will be comprised of 25 to 50 randomly selected citizens, who will meet for a minimum of three times over three months. The Citizens’ Council decides the number of participants and the length of the Citizens’ Panel.
In accordance with the legislation, the regional parliament is required to debate and respond to the recommendations developed by the Citizens’ Panels. The implementation of agreed upon recommendations is further monitored by the Citizens’ Council. The Ostbelgien model is the only example where this new institution extends the privilege of giving citizens a genuine voice in setting the policy agenda and providing citizens with the framework and tools to actively explore issues of their choice.
The remit of the first Citizens’ Panel is to explore options for improving the working conditions of healthcare workers. The Panel’s work began in spring 2020 but has been postponed as of the time of publication due to the global health crisis. The topic was chosen by the Citizens’ Council a few months before the new coronavirus (COVID-19) pandemic began.
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
City Observatory
The City Observatory is a model of a permanent deliberative body that was developed and implemented in Madrid city council in Spain. It was established through a regulation passed on 29 January 2019 that transformed the existing City Observatory into a representative deliberative body comprised of randomly selected citizens with new competencies. The initial Observatory was limited to a regular meeting between governing politicians and civil servants to analyse data about citizens’ opinions (collected through traditional means like opinion polls, focus groups, etc.). Until this regulatory change, the City Observatory had not held any meetings for several years.
The Observatory was designed by the Madrid city council, Participa Lab, newDemocracy Foundation, and other experts in the field of deliberation (Smith, 2019). However, after a change in government in May 2019, the future of the City Observatory was brought into question, and in February 2020, the composition and function of the Observatory reverted to a body comprised of governing politicians and civil servants.
Mandated to evaluate citizen proposals submitted through the decide.madrid digital participation platform, the City Observatory gathered 49 randomly selected citizens who met and deliberated over citizen proposals eight times per year (Madrid City Council, 2019). Observatory members had the power to send citizen proposals to a local referendum for a citizen vote, in this way opening up meaningful opportunities for participation both in setting the agenda and having a say in decision making. It is an example of digital democracy, deliberative democracy, and direct democracy being combined in an innovative way. The lessons drawn about institutionalising deliberative democracy in this way are discussed in Chapter 6.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Source: OECD Database of Representative Deliberative Processes and Institutions (2020).
Choosing a model of representative deliberative process
The most appropriate deliberative model depends primarily on the policy problem. The more complex the question is and the wider its implications, the more detailed recommendations are required and hence the more elaborate deliberative process is applicable. For example, Citizens’ Assemblies are well-suited to address constitutional questions and issues of national or greater importance, as this model allows for extensive learning about the policy issue and in-depth deliberation. They take place over 18.8 days spread over eleven months, on average. The participants, 90 on average, offer a large, diverse base of perspectives, which provides for greater legitimacy of their recommendations. However, they require more resources than any other model, due to their length and scale, as well as take much longer to set up and carry out.
Citizens’ Juries/Panels are focused processes to advise on a specific policy issue, typically at sub-national level, although they have also been used nationally/federally. As shorter, usually four-to-six day processes gathering 35-50 randomly selected citizens, they are long enough for citizens to develop detailed, informed recommendations to address specific policy issues, but require less time and less resources than Citizens’ Assemblies. They can thus be used more often and yield quicker results.
At the local and regional levels, a G1000 or a Citizens’ Council can be reasonable options for residents to develop a collective vision for a municipality and to address less complex community problems. As they can start from a clean sheet without a predetermined agenda, it is an opportunity to collectively identify and address the most pressing issues of the community or co-create a future vision. On the other hand, these options are less well-suited for in-depth recommendations on specific policy issues.
Another important consideration is how much flexibility should be given to participants during the deliberative process. More open-ended and flexible formats, such as the Austrian Citizens’ Councils that allow participants to shape the process, might lead to more out-of-the- box and creative ideas, and can be considered as well-suited to shape broader visions. The commissioning public authority needs to be open to genuinely considering all options that come out of such a process. However, if decision makers desire specific, informed recommendations for a pressing policy problem, then they need to clearly define the task for participants.
Other important considerations include available time and resources, level of government, and policy area. For example, the Consensus Conference model is helpful to assess technological advancements, as the format allows citizens to question scientists and policy makers extensively to get to the core of an issue. Figure 2.27 provides further indications on the properties of each model based on their use to date.
Overall, all models have trade-offs that need to be considered when making a choice. The longer the deliberative process is and the more participants it gathers, the more time and resources it will require. However, it is also more likely to result in more detailed and considered recommendations, and thus can be more useful for decision makers and more legitimate in the public’s eyes. The shorter a process is, the less well-thought through and detailed recommendations it is likely to provide, since it means there is less time for learning and deliberation. The stricter the design of the process, the less room there can be for creative ideas and solutions.
Source: Author’s own creation based on data in the OECD Database of Representative Deliberative Processes and Institutions (2020).
Combining features from different models
The process of choosing and tailoring the most appropriate representative deliberative model for a given context, level of government, phase of the policy cycle, and policy issue at hand is a creative one, with opportunities to combine features from different models. However, it is of essence to ensure that all fundamental phases of a representative deliberative process are preserved: learning, deliberation, and developing informed collective recommendations.
One practice is to combine several deliberative processes on different levels of government, which build on recommendations produced by preceding panels. An example of such a combination is a Policy Jury on School-based Clinics, conducted in Minnesota, United States in 1988. First, eight Citizens’ Juries, each comprised of 12 jurors, were set up in each of Minnesota’s Congressional districts to deliberate and provide recommendations on school-based clinics for the prevention of AIDS and teen pregnancy (Jefferson Center, 1988). They were then followed by a state-wide Citizens’ Jury, where three participants from each congressional district jury were gathered to bring recommendations together and decide on a final citizen stance.
Such a structure potentially allows for a greater number of citizens to be involved, to learn about the preferences of different regions, and to allow for informed deliberation between districts to reach a collective decision on a higher level of government. However, this type of combination also has potential disadvantages. Representatives from the process at the lower level of government will feel mandated by the outcomes established, undermining the deliberative nature of the process that takes place at a higher level of government since they will likely see themselves as representatives of the locality (Smith, 2019). In a deliberative process, participants need to be willing to consider the common interest and not come into the process with a strong mandate. These considerations should be taken into account when designing a process that combines citizen deliberation at different levels of government.
Some models also allow for some flexibility for the participants to shape the process. Predominantly an element of a G1000 or a Citizens’ Council, the opportunity to set the agenda, plan the proceedings, or decide how to self-organise during the deliberative process can and has been extended to other models, such as Citizens’ Juries and Citizens’ Assemblies.
Elements of deliberative models or a combination of several models have also been used to create permanent deliberative bodies. The most recent example as of early 2020 is the Ostbelgien Model, where a permanent Citizens’ Council is combined with the ongoing use of Citizens’ Panels. The result is a model that has features of agenda-setting, oversight, quality learning, deliberation, and informed recommendation development.
References
Asenbaum, Hans (2016), “Facilitating Inclusion: Austrian Wisdom Councils as Democratic Innovation between Consensus and Diversity”, Journal of Public Deliberation, 12(2), pp.1-11, https://www.publicdeliberation.net/jpd/vol12/iss2/art7, accessed 19 December 2019.
Breckon, J., Hopkins, A. and Rickey, B. (2019), “Evidence vs Democracy: How ‘mini-publics’ can traverse the gap between citizens, experts, and evidence”, The Alliance for Useful Evidence, https://www.alliance4usefulevidence.org/publication/evidence-vs-democracy/, accessed 9 December 2019.
Center For Deliberative Democracy (2019), “What is Deliberative Polling?”, Center For Deliberative Democracy, https://cdd.stanford.edu/what-is-deliberative-polling/ accessed 9 December 2019.
Center For Wise Democracy (2019), “Dynamic Facilitation”, Center For Wise Democracy, https://www.wisedemocracy.org/2-dynamic-facilitation.html, accessed 19 December 2019.
Escobar, O. and Elstub, S. (2017), “Forms of Mini-publics”, newDemocracy Foundation Research and Development Note, http://www.newdemocracy.com.au/forms-of-mini-publics, accessed 9 December 2019.
James S. Fishkin and Robert C. Luskin (1996), “The Deliberative Poll: A Reply to Our Critics”, in Public Perspective 7(1).
Foundation for Future Generations (2019), “German-speaking Community of Belgium becomes world’s first region with permanent citizen participation drafted by lot”, https://www.foundationfuturegenerations.org/files/documents/news/20190226_dgpermanentcitizensassembly_pressrelease.pdf, accessed 9 December 2019.
Fournier, Patrick, Henk van der Kolk, R. Kenneth Carty, André Blais, and Jonathan Rose (2011), When Citizens Decide: Lessons from Citizen Assemblies on Electoral Reform, Oxford: Oxford University Press.
G1000 (2019), “About G1000”, G1000.nu, available at: https://g1000.nu/over-g1000-2/, accessed 9 December 2019.
Healthy Democracy (2019), “Citizens’ Initiative Review”, Healthy Democracy, https://healthydemocracy.org/cir/, accessed 9 December 2019.
Jefferson Center (1988), “Final Report: Policy Jury on School-based Clinics”, Minnesota: Jefferson Center.
Madrid City Council (2019), “Observatorio de la Ciudad - Ayuntamiento de Madrid”, Madrid.es, https://www.madrid.es/portales/munimadrid/es/Inicio/El-Ayuntamiento/Observatorio-de-la-Ciudad/?vgnextfmt=default&vgnextchannel=38a9dec3c1fe7610VgnVCM2000001f4a900aRCRD&vgnextoid=38a9dec3c1fe7610VgnVCM2000001f4a900aRCRD, accessed 9 December 2019.
Nagano, Motoki (2020), “Citizen Deliberation Meetings (Shimin Tougikai), Proceedings of the International Symposium ‘Designing Deliberative Democracy: Practice and Experiments’” (18 & 19 February, 2020, Waseda University), Organiser: Planning committee of the International Symposium “Designing Deliberative Democracy: Practices and Experiments”, Center for Positive/Empirical Analysis of Political Economy, Top Global University Project of Waseda University / Co-organiser: Japan Mini-Publics Research Forum.
Nexus (2019), “Nexus Academy for Participative Methods”, Partizipative-methoden.de, https://www.partizipative-methoden.de/index.php/en/, accessed 9 December 2019.
Parliament of the German-speaking Community of Belgium (2019). “Was passiert beim Bürderdialog?”, Pdg.b, https://www.pdg.be/desktopdefault.aspx/tabid-5421/9372_read-56650, accessed 9 December 2019.
Partizipation.at. (2019), “Participation: The Citizens' Council”, https://www.partizipation.at/buergerinnenrat.html, accessed 9 December 2019.
Smith, Graham and Rosemary Bechler (2019), “Citizens Assembly: Towards a Politics of ‘Considered Judgement’ Part 2”, OpenDemocracy, https://www.opendemocracy.net/en/can-europe-make-it/citizens-assembly-towards-a-politics-of-considered-judgement-part-2/, accessed on 5 March 2020.
Smith, Graham (2019), “Institutionalizing deliberative mini-publics in Madrid City and German Speaking Belgium – the first steps”, Constitutionnet, http://constitutionnet.org/news/institutionalizing-deliberative-mini-publics-madrid-city-and-german-speaking-belgium-first, accessed on 10 April 2020.
Suiter, Jane (2018), “Deliberation in Action – Ireland’s Abortion Referendum”, Political Insight, September 2018: 30-32, https://journals.sagepub.com/doi/pdf/10.1177/2041905818796576, accessed on 14 November 2019.
The Danish Board of Technology (2006), “Teknologiradet”, Web.archive.org, https://web.archive.org/web/20081205112932 http://www.tekno.dk/subpage.php3?article=468&language=uk&category=12&toppic=kategori12, accessed 9 December 2019.
Warren, Mark and Hilary Pearse (2008), Designing Deliberative Democracy: The British Columbia Citizens' Assembly, Cambridge: Cambridge University Press.
World Wide Views (2019), “The World Wide Views method”, Wwviews.org, http://wwviews.org/the-world-wide-views-method/, accessed 9 December 2019.
Note
← 1. Lethbridge Citizens' Assembly on Councillor Employment and Compensation, Prince Edward County Citizens’ Assembly, Citizens' Assembly on Social Care, Camden's Citizens' Assembly on the Climate Crisis, and National Assembly for Wales
This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area. Extracts from publications may be subject to additional disclaimers, which are set out in the complete version of the publication, available at the link provided.
https://doi.org/10.1787/339306da-en
© OECD 2020
The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.
