Learnings from the French Citizens’ Convention on the End of Life

Kyle Redman

22.06.2023

4255 words, 18 min read

Marjan H. Ehsassi, Future of Democracy Fellow, The Berggruen Institute

mehsassi@berggruen.org

 

Helene Landemore, Professor of Political Science, Yale University & Research Fellow at Oxford Institute for Ethics in AI

helene.landemore@yale.edu

What was the challenge?

During his campaign for re-election, French President Emmanuel Macron made a commitment to open the debate on assisted suicide and euthanasia. For years, tension has been growing between those in favor of changing the law to introduce active assistance in dying and those who do not want France to move beyond the current law. 1 Under the 2016 Claeys-Leonetti law, terminally ill patients can request to be kept in a “deep, continuous sedation altering consciousness until death,” only if their condition causes “great suffering” and is likely to lead to a quick death. Euthanasia is currently legal in the Netherlands, Belgium, Luxembourg, and Spain under certain conditions. In Switzerland, assisted suicide is allowed where the patient takes a lethal dose of drugs themselves. Every year, French patients travel to these European countries in pursuit of options to end their life.

 

On September 13, 2022, President Macron announced the launch of a national debate to explore the possibility of legalizing end-of-life support. The Prime Minister, Elizabeth Borne, followed this announcement with a letter addressed to Thierry Beaudet, President of the Conseil Économique, Social et Environnemental (CESE), requesting a Citizens’ Convention on the End of Life (Convention Citoyenne sur la fin de vie, or CCFV) to engage with French citizens and to enrich the national debate on this subject.

A 2021 law adopted following the first French Citizens’ Convention for the Climate (CCC), empowered the CESE with greater authority. So far known as the chamber of organized civil society, it has been given a new mandate to engage with citizens and lead participatory practices – effectively creating a new institution for incorporating the informed common ground view of French citizens.

What happened?

185 French residents were chosen on the basis of randomly generated cell phone numbers, combined with statistical techniques of stratified random sampling, to serve on the CCFV. After that initial selection process was completed, only one person dropped out of the CCFV because they found employment after the first session and were no longer able to participate. The criteria used to select residents included gender, age, region, education level, urban/peri-urban/rural, and profession. Given the nature of the topic, some organizers advocated for the inclusion of religious and attitudinal criteria. In the end, because of the intrinsic sensitivity of asking about religious preferences in France, it was decided to proceed only with the original six criteria.

The government put a sufficiently open-ended question to the citizens, leaving open both the possibility of keeping the status quo or of modifying the law. A clear and unambiguous remit opened the door for targeted learning, strong deliberation, and an actionable final report:

Is the end-of-life support framework adapted to all situations or are changes necessary?

The 184 participants, broadly representative of the diversity of France, worked over 9 weekends, stretched over 27 days, at the Palais D’Iéna in the 16th arrondissement of Paris.  The 9 sessions were evenly divided by a Learning Phase, a Deliberation Phase, and a Harmonization Phase. On Sunday, April 2, the final day of the CCFV proceedings, the 184 participants presented their proposals to their fellow French citizens over livestream. The next day, on April 3, they were received at the Elysée by President Macron. Their final report along with their 65 proposals was approved by an overwhelming 92%. In the end, 76% of participants voted for a change in the law to introduce a French model for assisted suicide and euthanasia within a well-defined framework of strict guardrails.

What did we learn?

The CCFV will likely be recognized as one of the more impactful examples of citizens’ assemblies in recent times both for its process and expected outcomes. However, it was far from perfect. It is important that we draw lessons from this initiative as well as from other processes so that the design and implementation of future deliberative platforms can be refined to be more inclusive, responsive, and impactful. In the following sections, we share four areas that present a mix of strengths and weaknesses:

1. Governance Structure

What worked – Robust Leadership and Strong Working Relationships across Implementation Actors

The CCFV Governance Committee (Co-Gouv) was established by the CESE leadership. 2 Chaired by a senior CESE staff member, it consisted of fourteen individuals, six of whom were CESE employees (including the President of the Co-Gouv), and the other eight included experts in ethics, law, palliative care, and citizen participation as well as two former CCC participants. Additionally, A Collège des Garants (Committee of Guarantors) was set up and composed of four international experts in facilitation, philosophy, deliberation, and citizen participation to ensure compliance with principles of tolerance, pluralism, inclusion, and respect for the voice of all participants, in particular the minority voice and the marginalized.

From the outset, the President of the Governance Committee’s leadership style was open, direct, humble, and empathetic. The four guarantors of the Committee of Guarantors were included in all Governance Committee meetings that followed their appointment and were encouraged to provide perspectives on issues. The different implementation actors managed to successfully work together, despite disagreements and misunderstandings, by being committed to the same goal of an inclusive process and an impactful report. As trust grew, these factors set the stage for a process of decision-making that was generally congenial and committed to course correct when necessary.

What didn’t work – Centralized Decision-Making Process

The Governance Committee and other implementation partners worked well together; however, the decision-making process was highly centralized with the President of the Governance Committee, who is also a CESE member. For example, the welcoming speech delivered by the President of the Governance Committee to the citizens during the first session was written without seeking input from the 13 other members of the Governance Committee, despite requests to that effect and with problematic results, conveying the institutional CESE perspective on the role of citizens rather than the independent perspective of the Governance Committee itself. Several decisions were made by the President and CESE staff throughout the process without input from the full Governance Committee. Communication with the media was tightly controlled by the CESE, with the President as the only spokesperson for the Governance Committee. To create more balance on the Governance Committee and, crucially, preserve its independence as well as protect it from possible conflicts of interest, it might be advisable to either select a President who is independent from the CESE or select two co-Presidents, a member of the CESE and an external expert. In addition, having proportionally fewer CESE staff members on the Governance Committee would create some distance from the CESE and inject more external expertise into the day-to-day operations of the Governance Committee.

2. The Process, Facilitation and Final Report

What worked – Building Cohesion through Consensus and the Final Report’s Sections on Common Conviction and Manifest

From the outset, there was recognition that while the final deliverable was the product of citizens, it needed to be concise and actionable. There were discussions about whether limits should be placed on the number of recommendations and while the decision was to allow citizens to guide the final number of recommendations, an effort was made to streamline the large number of propositions with prioritizing votes. The CCFV process was designed to first engage around issues of convergence, namely palliative and hospice care. During the first sessions, working together to draft a framework around the report’s Common Conviction section helped build trust and create group cohesion. Later, when the CCFV broached more divisive issues, participants were able to listen to each other with respect and deliberate with civility. Juxtaposed with televised recordings of the National Assembly’s debates on retirement policy, the CCFV debates were calm and constructive.

Overall, the process of prioritizing and transcribing the recommendations on behalf of the participants was well managed by the facilitators who consistently worked late into the evenings. During the writing of the final report, the Harmonization Phase, Rapporteurs were selected from small groups to ensure that the work of participants was satisfactorily reflected at every stage. Another important contribution was the time set aside during the last session for the development of a Manifest. Through an innovative process of “vote without candidate,” six participants (three women and three men) were elected to produce the CCFV’s collective opening statement. This vote allowed citizens to select trusted peers for the challenging task of addressing the nation, which they did with great diligence and talent.  In the final analysis, the final report was well-constructed, concise, inclusive of all perspectives, and actionable.

What didn’t work – A Design Process that was not Sufficiently Planned and Shared in Advance, Inconsistent Facilitation and Quality of Deliberation

Consistent facilitation is a challenge in most citizens’ assemblies and the CCFV was not an exception. The core team consisted of facilitators from several organizations with Euro Group Consulting leading the effort. A comprehensive design manual was not shared with other implementing partners which made it difficult to understand the entire process from start to finish. Instead, every session was planned and described in between and sometimes during sessions. Some facilitators were outstanding, many were strong, and a minority was not sufficiently engaged or active. This was especially the case in ‘small group’ discussions which were in effect quite large and consisted of anywhere between 10-30 participants. Some participants reported frustrations around a lack of inclusion, the same voices being heard while others were ignored, and a failure to accurately incorporate their selection of words into the text. Small group discussions should aim for groups of 5 to allow for a share of voice and ownership of output. The inconsistency in facilitation was also evident in the quality of deliberation. During several sessions of the Deliberation Phase, small group discussions would have benefitted from more active facilitating techniques with kick-starter conversations to generate deeper analysis, reasoning, and stronger justifications for recommendations.

With respect to voting, deliberative platforms should rely on such processes thoughtfully and as circumstances demand. The CCFV decided to experiment with votes early and often, in part in reaction to the CCC which had only used votes during the last two session. The CCFV probably swung the pendulum too far in the other direction, with too many votes, too early and without a clear rationale for why preferences were being sought. There was also a lack of clarity on the wording of some votes as well as the process used, i.e. paper ballots and hand-counting. When the company Media Votes introduced its boxes, technical problems were encountered with the calculation of votes. An expert on the qualitative and quantitative dimensions of voting processes would have been an asset to the Governance Committee. The threshold was set at 66% for most votes and, in response to requests by participants, language was added to allow those against a change in the law to cast this vote each time (Contre L’Aide Active à Mourir). While cumbersome, this change allowed the minority voice to have a place on every ballot. Votes work well when clearly worded and serve one of two purposes: 1. Prioritizing: the vote is determinative in prioritizing the group’s position on an issue or a recommendation at a critical and mature point; and 2. Process Preference: the vote is on a process-related matter with participants voting on their preference as part of a co-responsibility or co-creation process. For instance, whether participants want their draft report to be reviewed by a fact-checking committee of experts.

3. Agenda Setting, Co-Responsibility, and Co-Creation

 What worked – Co-Creation Process Developed during the CCFV

During the first session, several guarantors, and members of the Co-Gouv raised the importance of creating an environment that empowered citizen leadership. CCFV participants also expressed a desire for co-responsibility and co-creation. They too wanted to be more involved in collective decision-making. A plan evolved; it was slow but steady. The Governance Committee instituted the practice of holding “open meetings” during the sessions to ensure a space for participants to provide feedback. After the initial sessions, four participants (two men and two women) were randomly drawn from the entire pool of participants to join the Sunday briefing session at the conclusion of the weekend. CCFV participants were encouraged to share their concerns with the volunteers so that their feedback could be shared at the meetings. The group of four was then invited to share their experience and insights with the larger group at the beginning of the next session. The President of the Governance Committee, held a press conference at the conclusion of each session. After session 6 and during session 7, approximately 40 participants signed a statement invoking their legitimacy and stating that they were better positioned to represent the voice of participants at CESE-organized press conferences. As a result, a decision was made to select two women and two men for each press conference. However, because the participants to the press conference were initially randomly selected from a pool of volunteers, some participants complained that they were represented by the usual suspects. As a result, the Co-Gouv decided to resort to the same selection process used to select participants to the briefing session: random selection from the entire pool of citizens, with the possibility of opting out. These measures were positive and helped shift more power to CCFV participants.

What didn’t work – Selection of Topic and a Co-creation Process that was Developed Reactively rather than Planned for Prior to CCFV

There was no initial CCFV plan for including the voice of participants in the decision-making process. In fact, a majority of the Governance Committee initially voted against including participants in the debrief sessions, even though that had been the practice of the previous Citizens’ Convention on Climate. However, as described above, the de facto inclusion of voices eventually occurred anyway through other measures and through votes put to the CCFV on questions of process. The guarantors only began their mandate during the first session, roughly two months after the Governance Committee began meeting and after the recruitment process was completed. Guarantors should be appointed and involved much earlier so that they can review and vet the design of a citizens’ assembly and share the process for greater transparency.

At a more general level, the French government may need to rethink the participation of the larger public in agenda-setting for citizens’ assemblies. So far it has not been possible for the public to be part of the agenda-setting. An increasing number of deliberative platforms are providing citizens with agenda-setting roles such as the selection of the topic. For instance, the Brussels Parliament will hold a deliberative committee in response to citizen-driven petitions. A Swiss national youth assembly is asking 18–24-year-olds to vote on their priority subject area. So far, mental health has received the greatest number of votes and will likely become the topic of the National youth assembly in the Fall. The selection of the topic should be driven by citizens. While French citizens were interested in the CCFV topic, they did not choose it. To be fair, a new petition system is currently being rolled out by the CESE, which should allow for that option in the near future. But the process to submit signatures is cumbersome and dissuasive, suggesting more work needs to be done to facilitate the use of this opportunity by the wider public.

Furthermore, to provide participants with a consequential voice and to create meaningful co-creation processes, elected representatives and practitioners should be guided by the following questions during the design and implementation of a citizens’ assembly:

  1. Have government champions of citizens’ assemblies been identified?
  2. Did citizens select the topic? Is the wording of the mandate/remit clear?
  3. Does the citizens’ assembly have a consultative or a binding decision-making mandate?
  4. Are roles and responsibilities and expectations clearly articulated and consistently communicated?
  5. Is sufficient time allocated for learning, deliberation, and the drafting of the deliverable? Has a diverse range of views been presented to the assembly? Have participants been able to request additional information and sources?
  6. Is the final report cohesive and actionable and is the writing by participants?
  7. Are all process-related details well thought-out, professionalized, and public?
  8. Is government present at the start and at the end of the process?
  9. Are follow-up mechanisms to ensure accountability clearly laid out by the government?
  10. What measures have been introduced for greater connection with the general public?

What worked – The CESE, the Manifest and the Minority Voice

During the reception at the Elysée, Macron announced his intention to engage with the CESE to plan future citizens’ conventions. As the third constitutional assembly of the French Republic, the CESE is well-positioned to lead innovative deliberation and consultation on pressing public problems on a national scale. In the future, the CESE’s growing legitimacy can accelerate the number of citizens’ assemblies across France and set up a funding process for ministries and government bodies to apply for grants to organize citizens’ assemblies at other levels of government (as a form of Democratic Action Fund) 3.

 

The minority voice was given considerable space, especially in the second phase of the process. After some citizens who found themselves in the minority complained to the Governance Committee and the guarantors that they did not feel heard, the Co-Gouv decided to introduce the principle of opinion groups in session 5 (out of 9). Remarkably these opinion groups did not polarize the larger group or created factions. Instead, the minorities opposed to assisted suicide grew more confident, vocal, and influential during both the working sessions and the plenaries, helping the majority opinions to become more understanding and informed. Trust between citizens grew as a result, as did the nuance and quality of the final document. This inclusion of minority provided another source of legitimacy for the CCFV. The initial sessions were designed to focus on areas of consensus so as to foster respect and social cohesion. A sense of community and trust emerged before moving into topics of divergence and disagreement.

 

While a few people continued to voice frustrations about not being sufficiently heard until the end, the broader sense of social cohesion was captured in the Common Conviction section of the final report and in participant testimonies and comments. The final report includes the arguments for and against the propositions to ensure that despite differences, all voices were adequately represented. During the last session, one of the most vocal leaders of the minority against assisted suicide and a change in the current law thanked the 76% majority for providing the minority with 50% of the speaking time and 50% of space in the final report. The Manifest was addressed to the public and the last paragraph states that: “our completed work is now yours. It is the result of collective deliberation, the conclusions of which we bring to the public debate.” 5

What didn’t work – Inconsistent Government Messaging and an Unengaged Broader Public

Several elected representatives participated in the CCFV process. During the first session, The Prime Minister, Elisabeth Borne, addressed the Citizens’ Convention, along with the President of the Assemblée Nationale, Yaël Braun-Pivet. After several questions about the government’s commitment to act on the CCFV’s recommendations, Braun-Pivet became increasingly frustrated and told participants that they must not forget that they, unlike elected representatives, do not represent French people and do not have the legitimacy to make decisions. This caused a stir in the audience. The President of the CESE, Thierry Beaudet responded by stressing the Convention’s legitimacy and reminding everyone that in fact the CCFV was more diverse and therefore more representative of France than any elected body (playing on the ambiguity of different types of representation). Later, Olivier Véran, Minister Delegate for Relations with Parliament and Citizen Participation, joined the proceedings as well as Minister of Health, Agnès Firmin-Le Bodo. Given the initial letter from Prime Minister Borne, there was an expectation that the CCFV would submit their final report to Borne at the closing session. The Prime Minister was not able to join the participants at the Palais D’Iéna. Instead, they released their report online over livestream to the French public and delivered it in person to President Macron at the Elysée Palace on April 3, 2023.

Finally, reaching the broader public is an important component of building legitimacy for a citizens’ assembly. The more familiar the public is with a citizens’ assembly, the more the citizens’ assembly is perceived as representative of the larger community.  A decision was made early in the process not to have an online platform to collect input from the broader public (largely for reasons of manageability). From the outset, some CCFV plenary sessions were livestreamed. After the Learning Phase, the public could register to attend sessions in person. The press conferences with participants also helped increase knowledge of the CCFV. Initially, the participants did not want the media to cover their small group discussions. However, over time, they felt a sense of responsibility to share the process and their work with the French public. The media was then invited to attend smaller group discussions which helped to raise more awareness. However, a greater focus on the profiles of the participants would have been beneficial. For instance, in 2019, the New York Times published an entire section that brought the 526 randomly selected American citizens who made up America in One Room into the homes of the American public. More events on Zoom, as well as an online public education campaign like the one set up by Belgium’s G1000 We Need To Talk would have created a greater connection between the CCFV and the general public. In general, the link between deliberative processes and the public continues to be weak. Governments and the deliberative community should engage in deeper reflection about the architecture of citizens’ assemblies to strengthen this connection and create more meaningful points of engagement between the two.

Next Steps

President Macron requested that the National Assembly build on the recommendations of the CCFV and introduce a projet de loi – a legislative bill – by the end of Summer 2023.  The CESE has committed to a follow-up session with CCFV participants within six months. A formal recommendation from the CESE Committee on the End of Life is to be released in the following weeks. It is expected that this report will complement the work of the CCFV. The participants have also considered establishing Les 184, an association to help guide their work in the months ahead. Finally, discussions are ongoing for additional CESE-led citizens’ conventions on other issues of national importance.

CCFV policy outcomes have yet to be fully realized and there have been mixed reviews of Macron’s participative legacy (critics have called it a form of “participatory authoritarianism”), however some preliminary observations can be made. We often speak of power without deliberation or deliberation without power. The Citizens’ Convention on the End of Life demonstrates that with strategic leadership and clear objectives, a citizens’ assembly can combine deliberation and power. With political will, a consultative mandate need not be binding to be consequential. Given CCFV’s success in process and outcome thus far, it will likely become a strong standard bearer for the next wave of citizens’ assemblies worldwide.

Suggestions for improvement

To increase their impact, future citizens assemblies should refine their process design and implementation with a focus on:

  • A balanced governance structure that is more independent of the CESE and shares power with participants through a robust agenda-setting and co-creation process.
  • Better planning, anticipation, and communication of every step to the citizens
  • A more open flow of information with a process that is shared from the outset to improve transparency, and lead to better planning and greater fluidity between sessions.
  • Enhanced training in deliberative facilitation and more consistent facilitation to encourage more voices, improve participant experience, the quality of deliberation and of the final report.
  • Stronger connection with the wider public to increase recognition, enhance legitimacy and credibility of the process.