Not Just One Thing: Citizens’ Assemblies and Citizens’ Panels After Twenty Years of Democratic Experimentation with Minipublics

This paper examines the twenty-year expansion of minipublics and argues for a distinction between two models, the citizens’ assembly model and the citizens’ panel model, only the first of which can make a strong representative claim. Although contemporary discourse often gives the label “citizens’ assembly” to both, we will reserve that term for minipublics that have over one hundred members, deliberate for more than ten days, are open to all citizens on the electoral roll, incorporate recursive representation, and have their own regulation to ensure they operate at arm’s length from the convening body and special interests. By contrast, we will call minipublics “citizens’ panels” when they have fewer than 100 members, deliberate for fewer than 10 days, have relatively flexible criteria in sortition, lack strong mechanisms of recursive representation, and do not have the regulatory safeguards needed to ensure full independence. Minipublics have evolved into these two distinct forms not by chance, but for normatively meaningful reasons. Citizens’ assemblies exist primarily to represent citizens—to speak or decide on their behalf. Citizens’ panels are designed primarily to getting citizen input from less frequently heard voices. Citizens’ assemblies, as we define them, make a strong representative claim. Citizens’ panels make a weaker claim to political representation but nonetheless serve important purposes in eliciting citizen input. We depart from current terminology. At present, both in theory and practice, these two distinct models are often grouped together under the single label of “citizens’ assemblies.” We also depart partially from current academic typologies. Normally, typologies in the field cover an array of categories with a large-format one, usually with the label “citizens’ assemblies,” a smaller one, with the name of “panels” or “juries,” and five or six more categories covering, for example, deliberative polls, citizens’ initiative reviews, planning cells, and consensus conferences (Curato et al., 2021; Escobar & Elstub, 2017, 2025b; Harris, 2019; Ryan & Smith, 2014). We suggest simplifying these categorizations to only three main categories: citizens’ assemblies for large-format minipublics with a strong representative claim; citizens’ panels with a weak representative claim; and deliberative polls, which have a third function of conveying considered public opinion. Citizens’ panels are the broader category of minipublics, encompassing many different formats, including citizens’ initiative reviews, planning cells, and consensus conferences. Rather than proceeding deductively, this survey takes an inductive approach—drawing lessons from practice rather than imposing them from theory. It also aims to be global: by analyzing not only the prominent cases, which are primarily in the Global North, but also smaller, often overlooked processes, many of them in the Global South. We have paid particular attention to including recent examples of minipublics, drawing especially on experiences from 2024 and 2025. The article is a mixture of descriptive analysis (what is), normative analysis (what should be and why), and practical suggestion to the deliberative community.