Deliberative mini-publics are gaining traction globally to address complex issues. Citizens’ assemblies, a specific type, involve randomly selected citizens deliberating and producing policy recommendations. Climate action is a prominent issue, and the French Citizens’ Convention for Climate (CCC) is a large-scale example. The CCC, initiated in 2019, tasked 150 citizens with proposing measures to reduce greenhouse gas emissions by at least 40% by 2030, emphasizing social justice. A key feature was its “innovative co-construction procedure,” lacking clear definition. This raises questions: if external actors are involved, who is responsible for the outcome? Can co-construction succeed without broader public resonance? This paper examines these questions by analyzing the CCC proceedings using qualitative and quantitative data gathered by a research team.

The paper reviews literature on citizens’ assemblies and co-construction. Citizens’ assemblies aim to enhance democratic quality by complementing representative democracy through inclusive deliberation. While there's evidence of mini-public effectiveness in generating public support, the degree of consequentialism remains debated. Climate action is a crucial focus, as deliberative democracy is better suited to handling complex, intangible issues like climate change. However, the role of expert input needs careful consideration to maintain citizen agency. Co-construction, a broader concept than deliberative democracy, involves citizens and state representatives in policy-making. While sharing the goal of enhanced democratic quality, co-construction differs from citizens’ assemblies in terms of citizen primacy and the involvement of the broader public. The paper seeks to address the ambiguity of responsibilities and the role of non-participants in a co-constructive citizens' assembly.

The research employed a mixed-methods approach, involving a team of 30 social scientists who observed CCC proceedings, collected qualitative and quantitative data, and conducted surveys. Observation included plenary and group discussions, with researchers maintaining a non-interfering role. An internal survey gauged citizens' values, attitudes, and perceptions, while an external survey tracked public opinion. Quantitative data included voting data and information from the CCC's online platform. The study used a structure-process-outcome framework to analyze the CCC proceedings and their context, considering the interactions between citizens and steering bodies at key stages: agenda-setting, proposal development, and decision-making. The research also examined the interactions between the citizens and the broader public.

The CCC, while initially aiming for a 40% reduction in greenhouse gas emissions by 2030, faced disruptions from protests and the COVID-19 pandemic. Despite this, 149 measures were approved. Analysis reveals that steering bodies, particularly the governance committee, had significant influence, but citizens retained agency, evidenced by their rejection of the carbon tax issue and adjustments to the agenda. Expert input, particularly from the legal advisory group, was crucial in refining proposals, but did not fully diminish citizen creativity. While high consensus emerged among participants, the co-constructive approach failed to generate substantial public support. The citizens, focused on internal consistency, largely ignored public input. They also rejected a referendum on most measures, prioritizing their perceived credibility and avoiding involvement in public campaigning. Post-assembly evaluation showed citizen disappointment with the government's follow-up, despite generally positive views on citizens' assemblies. The study reveals that the co-constructive approach of the CCC fostered closer relations between citizens and policymakers but failed to connect citizens with the broader public. The high approval rates for the measures can be partly attributed to voting by blocks and the lack of a measure-by-measure vote.

The findings challenge the assumption that citizens’ assemblies automatically generate public support. The CCC's success in internal consensus building did not translate to broader public acceptance. The lack of a clear commitment structure regarding government follow-up contributed to this disconnect. The citizens’ decision against a referendum highlights potential tensions between mini-publics and broader public input. It also suggests that a citizens' assembly may be perceived, by its own members, as a legitimate process for generating recommendations, but not a sufficient means for establishing broader political legitimacy for the adopted measures. The study emphasizes the importance of a clear commitment structure to enhance consequentiality and legitimacy in citizens’ assemblies.

The CCC demonstrates the potential of citizens’ assemblies to improve deliberation quality, particularly with expert support. However, it highlights the need for clear commitment structures to ensure that the outcomes resonate with the wider public and are effectively implemented by the government. Future research should explore ways to improve communication and engagement between citizens’ assemblies and the broader public, addressing the challenge of securing public support for complex policy issues.

The study acknowledges limitations in data collection due to logistical challenges (e.g., recording group discussions) and survey response rates. The non-interfering observation approach might have limited the depth of interaction with participants. The study also notes that the lack of a pilot test for the internal survey led to an unintended bias in some questions. The context of significant national events also significantly affected the process, but it might not affect the generalizability of the findings.