Designing deliberative democracy: A global conjoint experiment on citizens' assemblies

Citizens' assemblies—institutions that provide a space for normal citizens to engage in deliberation on public policies choices—have been lauded for their potential to enhance democracy and to increase the buy-in for unpopular measures, including in the realm of climate change. This paper will present results from a survey experiment (N=41,428) conducted in 18 EU countries and India, the United States, Brazil, Japan and the United Kingdom. The survey experiment tested the public’s preferences regarding six design attributes of citizens’ assemblies: (i) who should participate; (ii) how many participants there should be; (iii) the decision making mechanism, (iv) the role of the recommendations; (v) access of the public to the deliberations; and (vi) whether deliberations should take place in-person or online. The respondents were presented with, and asked to rate, five randomly selected pairs of design features of a citizens’ assembly, intended to involve their country’s public in the process of decision making on climate policies. We find there is a clear preference for holding a referendum on the recommendations of a citizens’ assembly in which all citizens can vote rather than implementing the policy suggestion directly or using them to inform the decisions by parliamentarians. A clear majority is in favour of making the deliberations open to the public. Overall respondents prefer larger assemblies. Respondents also prefer all citizens willing to take part to be enabled to do so rather than selection through sortition. The paper will draw out implications for deliberative democracy scholarship and practice. It will discuss the feasibility of these recommendations and the lessons learned from former citizens’ assemblies for the design of future deliberative democracy initiatives.