Multi-body sortition

A single citizens’ assembly does one job at a time. Multi-body sortition asks a bigger question: what would an entire legislature look like if it ran on sortition instead of elections? The fullest answer is a design by Terry Bouricius — a former Vermont city councillor and state legislator turned sortition theorist — set out in his 2026 book Democracy Without Politicians. Rather than one body that debates, drafts, and decides, he splits those functions across several randomly selected bodies, each with its own size, term, and selection method tuned to its task. The separation is the point: no single group both writes a law and passes it, which is what builds in accountability and blocks the capture that one all-powerful chamber would invite.

The bodies

Bouricius calls it a “reference design,” like an architect’s — a complete blueprint from which real jurisdictions would use some parts and not others (a city might run Review Panels and Policy Juries in one policy area while keeping its elected council for everything else).

• Agenda Council (~150 members) — sets which topics the coming period will address and issues a call for proposals, in place of politicians picking hot-button or lobbyist-friendly issues. An optional petition route lets the public add topics.
• Interest Panels (~a dozen each, as many as needed) — open, often self-selected groups that draft proposals on the agenda topics. Because they only raise ideas and never decide, opening the gates wide is safe: an unelectable expert can contribute, and a bad proposal simply fails at the next stage.
• Review Panels (one per policy domain) — stratified-random bodies, full-time at national scale, that do the committee work: hearings, expert witnesses, professional staff, combining and amending the Interest Panels’ drafts into a finished bill. They craft legislation but, deliberately, never pass it.
• Policy Juries (500–1,500 members) — large, near-fully-representative bodies that make the final call. Like the Athenian nomothetai, they hear balanced pro-and-con presentations and then vote by secret ballot without debate — leaning on the wisdom of crowds while sidestepping the groupthink and pride-of-authorship that discussion can breed. A law passes only if an informed cross-section of the public approves it.
• Coordination Council — harmonises policies across the siloed Review Panels (budgets especially); its proposals also face a Policy Jury.

Two meta-legislative bodies govern the system itself:

• Rules Council — a lottery body that tunes the rules (the lottery method, the size and term of each body, how expert witnesses are chosen). Its changes must clear a Policy Jury, so it cannot expand its own power. The rationale is Rawlsian: a body that will never itself pass a law is best placed to set fair rules for the bodies that do, because it decides behind a veil of ignorance. As Bouricius puts it, “it is the having of power to adopt laws that can corrupt — not merely whether the representatives are selected by election or lottery.”
• Oversight Councils — lottery bodies that police staff performance and fairness rather than policy: they rule on complaints about biased presentations and can hire and fire the staff serving the other bodies.

Why split it up

The design answers a standard worry about sortition — that a single random assembly could be steered by its staff, captured, or fall into groupthink — by refusing to concentrate the functions in one place. Each body is too limited to be worth capturing, and each checks the others: Interest Panels propose but don’t decide, Review Panels draft but don’t pass, Policy Juries decide but don’t draft, the Rules Council writes the rules but can’t adopt laws. It is a separation of powers built inside a sortition system rather than between elected branches. Truncated versions already exist — the Ostbelgien standing council and Paris’s permanent 100-member citizen body pair an agenda-setting council with deliberative panels — which is why Bouricius treats the full blueprint as a direction of travel, not a finished constitution. It is the people’s branch idea taken to its conclusion: not a chamber bolted alongside elections, but a legislature without politicians.

Accountability without elections

The sharpest objection to any sortition system is that you can’t vote a random body out — so how is it accountable? The answer is that elections barely deliver accountability anyway (incumbents are re-elected above 90%, and voters can’t realistically monitor what representatives do), and that the multi-body design supplies a sturdier kind. Three mechanisms do the work. First, congruent accountability: because each body is a microcosm of the public — John Adams’s “exact portrait of the people at large” — it tends to decide as the public itself would, so accountability is built in by representativeness rather than bolted on as the threat of removal. Second, separation of tasks: no body is “a judge in its own cause” (Madison’s phrase), so corrupting an outcome means corrupting several independent bodies at once. Third, rotation and short, single-issue juries: power can’t accrue where terms are brief and members never face re-election, campaign fundraising is absent, and votes are cast by secret ballot with anti-tampering protections (Bouricius would even seed the process with pretend bribers who reward those that report them, echoing Athens’s audit of its magistrates). Because each jury is fresh, the system is self-correcting — a bad decision carries no pride of authorship and tends to be reversed by the next jury rather than defended for years. Bodies are still expected to “give an account,” publishing the reasoning behind a decision. The claim is not that a sortition democracy needs no anti-corruption rules, but that it starts with a large head start over an elective one — see objections to sortition for the fuller exchange.

Getting there: the peeling strategy

How would a system like this ever replace an elected legislature? Bouricius rejects the two ideas that occur first. A sortition second chamber (an allotted house alongside an elected one — the tempting fate for the UK’s House of Lords) and mixed-member bodies (some seats elected, some allotted) both put random citizens head-to-head with career politicians over the same bills — and the politicians, practised communicators with careers to defend, will win. When the two clash they attack the lottery body’s legitimacy (“a random gaggle of dishwashers and hairdressers … unaccountable, because they never have to face you in an election”), while the partisan agenda bleeds through and pulls the sortition side back into team loyalty. The Brussels Parliament’s 2019 Deliberative Committees — lay citizens seated with MPs — bore this out: researchers found “power asymmetries clearly remained,” many legislators didn’t attend, and only the politicians’ votes ultimately counted. Begun this way, Bouricius argues, sortition is “more likely to die there than expand.”

His alternative is the peeling strategy: rather than contest the same bills, peel away one issue area or function at a time from the elected legislature and hand it to a sortition process — pensions this year, say; city planning the next. Because the multi-body design is already a network of single-issue, single-function mini-publics, the transition is built into the model: domains can be transferred to the sortition wing of government one at a time, letting communities learn the craft incrementally and avoiding the head-on collision that gets sortition killed. Elections and sortition coexist through the handover (and could indefinitely), but never as rivals in the same chamber.

Sources

• Terrill Bouricius, Democracy Without Politicians (Routledge, 2026) — the multi-body “reference design” (ch. 16), the accountability argument (ch. 13), and the “peeling” transition strategy (ch. 17).