Skip to content
Citizen Infrastructure

What is a citizens' assembly

A citizens’ assembly is a group of ordinary people, chosen by lot to mirror the wider population, who come together to learn about an issue, deliberate, and make recommendations. It’s the format behind most of the deliberations in our Run Reports — from the Irish assembly on the Eighth Amendment to the French convention on end of life.

How it works

Section titled “How it works”
  • Selection by lot, stratified. Members are picked at random, but with quotas — typically gender, age, education, socioeconomic background, and geography — so the assembly is a fair cross-section of the public (a “mini-public”). That cross-sectional quality is what gives it legitimacy: it looks and thinks like everyone else. Choosing which categories to balance is itself a judgement — the French end-of-life convention, for instance, debated whether to include religion and decided against it.
  • Learn, then deliberate. Members hear balanced expert evidence and a range of viewpoints, then deliberate in small facilitated groups over multiple sessions — often several weekends or longer.
  • Experts inform; citizens decide. A governance or steering committee usually sets up the process (duration, criteria, who’s invited). In the strongest designs, once it’s running the experts step back and the citizens make the calls.
  • Recommend. The assembly produces recommendations — to a government, a parliament, or a referendum.

What it’s good for (and not)

Section titled “What it’s good for (and not)”

Because members aren’t defending fixed party positions, assemblies tend to treat hard questions as shared problems, which often lets them reach considered, broadly-supported conclusions on issues that paralyse ordinary politics. Their weakness is usually the last mile: most are advisory, so their recommendations can be filtered, diluted, or ignored by the institutions meant to act on them — which is why some, like the Ostbelgien model, are being made permanent and given real standing.

What impact do they have?

Section titled “What impact do they have?”

Beyond producing recommendations, an assembly can change three quite different things — and the evidence base for each is very different. People Powered’s 2025 evidence brief, synthesising 66 empirical studies across five continents, sets out three distinct levels of impact: personal (what changes for the members themselves), spillover (what changes for the wider community), and institutional (what changes in the government or agency the assembly addressed). Personal impact has the strongest evidence base; institutional impact the weakest. Judging any given assembly is easier once you know which level you’re asking about.

How to tell if it was done well

Section titled “How to tell if it was done well”

Beyond the general craft of running an assembly, practitioners at FIDE North America use a compact seven-component evaluation framework — a checklist for whether an assembly hit the bar:

  1. Representativeness — did the sortition draw actually mirror the population, and were the stratification categories chosen honestly?
  2. A clear remit — was the question specific enough to be answered?
  3. Time — did members have enough of it to learn, deliberate, and revise?
  4. Diverse information — was the evidence balanced, with dissent given real airtime?
  5. Deliberation — was there genuine facilitated small-group discussion, not just voting between speakers?
  6. Free response — could members steer the discussion and change their minds, or were they steered toward a pre-cooked answer?
  7. Accountability — did the commissioner commit to a public, point-by-point response and follow through?

The list is deceptively unglamorous but load-bearing: skip any one component and the assembly tends to fall over on that dimension, in ways members can feel even when outside observers cannot.

Sources

Section titled “Sources”