The Role of Global Citizen Assemblies in Civilizational-Scale Deliberation

Every major civilizational challenge of the 21st century shares a structural feature: the decisions that matter are made in institutions whose design predates the problem, whose geographic scope is narrower than the challenge, and whose accountability mechanisms face primarily the people they were designed to serve rather than the people most affected by their choices.

The United Nations Security Council was designed in 1945 around the power relationships of the post-World War II moment. Five states hold permanent vetoes. The logic was realpolitik: make the agreement acceptable to the most powerful states by guaranteeing their interests. The result is a body structurally incapable of producing decisions that any of its five permanent members opposes, which means it is structurally incapable of governing any problem where those members have divergent interests — which describes nearly every serious civilizational challenge.

National parliaments and congresses are accountable to their electorates, which means they face systematic pressure to prioritize national interests over global ones, short-term costs over long-term benefits, and the preferences of current voters over the interests of future generations or non-citizens. This is not a failure of politicians — it is the designed operation of the system. Representative democracy at the national scale optimizes for national preference aggregation. It does not optimize for civilizational revision.

International negotiating bodies — the Conference of Parties processes, the World Trade Organization, the various UN specialized agencies — operate by state consensus, which means every revision to existing arrangements must be acceptable to every state that wishes to block it. The result is the phenomenon familiar from climate negotiations: agreements calibrated to the lowest common denominator of state willingness, systematically underambitious relative to what the problem requires, and routinely undermined by non-binding commitment structures.

This is the governance gap. The problems are civilizational. The institutions are national and consensual. The revision mechanisms are slow, narrowly accountable, and systematically exposed to vetoes by actors whose interests are threatened by the revision.

The citizen assembly model emerged from political philosophy's engagement with deliberative democracy — the idea that legitimate political decisions should be the product of reason-giving deliberation among free and equal citizens, not merely the aggregation of pre-formed preferences. The practical implementation was developed through a series of experiments beginning in British Columbia in 2004, where a Citizens' Assembly on Electoral Reform was convened to evaluate the province's voting system.

The model has several defining features:

Sortition: Participants are selected by stratified random sampling to achieve demographic representativeness across relevant dimensions — age, gender, geography, education, socioeconomic status. This is the same method used for jury selection. The randomness removes the selection biases inherent in electoral politics: you cannot run a campaign to get onto a citizen assembly; you can only be selected. This means the assembly is not dominated by the politically engaged, the ideologically motivated, or those with resources to seek office.

Learning phase: Before deliberating, assembly members receive structured information. They hear from scientific experts, affected stakeholders, advocacy groups on multiple sides, and practitioners with implementation experience. Crucially, they are supported in understanding disagreements among experts — they are not presented with a false consensus that papers over genuine uncertainty. This phase typically runs for weeks or months on complex topics.

Deliberation: Small-group discussions, facilitated by trained moderators, give all members substantive voice. The facilitation structure is designed to prevent domination by confident or high-status participants. Members are supported in articulating their reasoning, not just their conclusions.

Recommendation: The assembly produces a recommendation — typically with a vote showing the degree of consensus — along with a written account of the reasoning behind it. The recommendation is usually not legally binding, though in some cases it has been directly linked to legislative process or referendum.

Evidence base: Citizen assemblies on climate have consistently produced recommendations more ambitious than those produced by parliamentary bodies operating through normal channels. The French Citizens' Convention for Climate, convened in 2019-2020, produced 149 proposals, several of which were directly incorporated into French climate legislation. The Irish Citizens' Assembly on biodiversity and climate produced recommendations that moved significantly beyond existing policy. The UK Climate Assembly produced nuanced recommendations that distinguished citizen willingness to support certain measures from willingness to support others in ways that polling alone could not capture.

The pattern across multiple contexts is consistent: when ordinary people are given the time, the information, and the facilitated space to deliberate on complex problems, they are capable of producing recommendations that are more nuanced, more ambitious, and more sensitive to trade-offs than the outputs of normal representative processes. This is not because citizens are smarter than politicians. It is because the conditions of deliberation are different: no electoral accountability to a base, no party discipline, no donor relationships, no reelection calculus. The assembly is accountable only to the quality of its reasoning.

The case for extending the citizen assembly model to the global scale follows directly from the governance gap analysis. If national assemblies can deliberate on national climate policy with unexpected quality, a global assembly could deliberate on global climate governance with the additional advantage of including perspectives from populations currently excluded from the state-centric negotiating process.

The obstacles are substantial and should be taken seriously rather than wished away.

Language and translation: A genuinely global assembly requires real-time, high-quality interpretation across dozens or hundreds of languages. This is logistically demanding and expensive, but it is a solved problem. The UN operates routine multilingual deliberation. The question is resourcing, not impossibility.

Digital access and exclusion: Globally representative sortition requires participant selection from populations with highly variable digital infrastructure. Remote participation — the model used by most post-2020 assemblies — excludes populations with unreliable internet access, which correlates precisely with the populations already most excluded from global governance. Hybrid models combining in-person and remote participation can partially address this, but design choices matter and will reflect political priorities.

Geographic and temporal coordination: Coordinating deliberation across time zones, with culturally appropriate facilitation, across widely varying political and social contexts is genuinely hard. It requires substantial investment in facilitation capacity, translation, and logistical support. It is not a reason to abandon the concept; it is a design challenge.

Authority and binding weight: The most significant obstacle is not logistical but political. A global citizen assembly has no formal authority over any state. Its recommendations are moral pressure, not legal obligation. This is the same limitation that applies to most international agreements — states comply when they choose to, for reasons related to self-interest, reputation, and domestic political pressure. The question is whether a well-conducted global assembly, with genuine representativeness and high-quality deliberation, can generate the kind of moral pressure that moves state behavior.

The COP26 Global Assembly in 2021 involved 100 participants from across the world, deliberating for three weekends on climate governance. It was too small to be genuinely representative, too resource-constrained to be fully supported, and too disconnected from official processes to influence them directly. But it demonstrated that global deliberation across language and geography is achievable, that participants engage seriously with complex material, and that diverse perspectives can produce coherent shared recommendations. It is a proof-of-concept, not a mature institution.

The deeper contribution of citizen assemblies to civilizational revision is not procedural but epistemic. They address a specific failure mode in institutional decision-making: the systematic exclusion of non-expert, non-elite, non-represented perspective from the generation of solutions.

Technical expertise is essential to governing complex systems. But expert communities develop shared assumptions, shared blind spots, and shared interests that can distort their recommendations in ways that are invisible from inside the community. Climate scientists agree on the physical reality of climate change with overwhelming consensus; they have been far less unanimous on questions of justice, distribution of adaptation costs, and the political feasibility of proposed solutions. These are not purely technical questions. They require input from the people who will live with the answers.

State negotiators represent their governments' interests. Those interests include the welfare of their populations but also the interests of incumbent industries, geopolitical alliances, and domestic political coalitions. A state negotiator cannot easily break from a national interest position, even if they privately believe it is inadequate to the global problem. The citizen assembly model creates a space where individuals, temporarily freed from institutional accountability, can reason together without those constraints.

The output of a global citizen assembly functions as a reference point — a statement of what ordinary people, given the information and time to reason carefully, actually conclude. This reference point can be used by advocates inside negotiating processes to pressure their own governments. It can be used to build public awareness of where citizens stand relative to where governments are positioned. It can be used to legitimate more ambitious action by governments willing to use it.

Law 5 — Revise — holds that systems that cannot revise themselves accumulate error until they fail catastrophically. The governance architecture currently managing civilizational-scale problems is not revising itself at the rate the problems require. It is producing incremental adjustments to existing positions, constrained by the interests of the most powerful actors in a system designed around those interests.

Global citizen assemblies are one mechanism — not the only mechanism, not a sufficient mechanism, but a genuinely useful one — for introducing new inputs into that system. They bring perspectives from populations currently excluded from formal deliberation. They surface the considered judgments of ordinary people across diverse contexts, which diverge in important ways from the positions of their governments. They create accountability pressure from below and from outside the normal negotiating channels.

They are not a replacement for state authority, technical expertise, or formal international law. They are a complement — a way of enriching the quality of civilizational deliberation by making it genuinely more representative of the civilization it is supposed to govern.

A civilization that can revise its governance architecture to include the considered judgment of its ordinary members, even on the most complex and globally consequential questions, is a civilization that has not given up on the possibility of democratic self-governance at scale. That is not a small thing. In the current historical moment, it may be one of the most important institutional experiments underway.