The rule, exactly. Before an election each citizen chooses: to vote — or to take a payment and not vote in that election. The payment is a percentage of the median income, set by referendum; there is no fixed sum. One ballot is one vote, with no multiplier in the count; the vote is amplified only in that each ballot's share grows as others step out. The budget pays under law; a candidate never pays. Only a referendum of all citizens — simple majority, no quorum — introduces, changes or repeals the rule.
The protocol has been introduced nowhere and no pilot has been run: shares, turnout and outcome figures in the chapters are estimates, and the protocol promises nobody an election victory. If a chapter says otherwise, Exact Answers and the Charter are correct. For a candidate: ten questions and ten steps. For a citizen, a mayor, a finance officer, a donor, a journalist, a scholar, a lawyer: answers by role. Everything in force in one file: llms-full.txt.
23e. Deliberation and Citizens' Assemblies: Quality Without a Door¶
Chapter: 05 — Empirical Base File: 05_023e · v1 · 30 September 2026 (session 30.09.26) Source: the architect's decision "gather everything already in the repository into separate chapters"; consolidated from 019 §61.2 and §61.5, 023 §23.2–23.3, 036 §85.2 and §85.5; figures checked on 30.09.2026.
1. Why a separate chapter¶
Deliberation is the only reform among those compared whose effect has been measured on the quality of a decision rather than on turnout. The repository mentions it in four places, each time in a line. It is brought together here because it is the best available argument for the protocol's main assumption: a small group of people who care about a question decides better than a large one that does not.
2. What happened¶
Deliberative polls. James Fishkin proposed the method in 1988: a random sample of citizens is gathered, given materials and experts from different sides, and opinions are measured before and after. More than a hundred such polls have been held in thirty countries. The result is stable: knowledge increases, extreme positions soften, and on average about two-thirds of the measured opinions change.
Ireland. A Citizens' Assembly — 99 people chosen by lot and chaired by a judge — sat from October 2016 to April 2018; five of its twelve meetings were devoted to abortion. 64 % of the members favoured lifting the restrictions. At the referendum of 25 May 2018 the amendment was adopted by 66 % of voters: the assembly's conclusion and the country's vote coincided almost exactly.
France. The Citizens' Convention on Climate adopted 149 proposals in June 2020. The president promised to support 146; by the official count a hundred of them were later implemented in full or in part — estimates of how "partial" differ.
Deliberation Day. In 2004 Ackerman and Fishkin proposed doing the same for a whole country: a holiday two weeks before an election and 150 dollars to everyone who comes to deliberate. It has been adopted nowhere.
3. What the repository already says¶
- The effect on quality is plus 5–10 % locally; on turnout and on populism, zero. It is an instrument for particular decisions, not an electoral reform (023 §23.2–23.3).
- The recommendations bind nobody: a parliament may emasculate them (019 §61.2, 036 §85.2).
- A vision without a door: there is no mechanism by which deliberation would become a general rule, because it must be introduced by those it constrains (019 §61.3).
- Compatibility is high: the protocol first removes the uninterested, and deliberation begins to work better, because those who remain to deliberate are the ones who care (036 §85.5).
4. The comparison¶
| An assembly, a deliberative poll | The protocol | |
|---|---|---|
| Who decides | a small group by lot | everyone who stayed by his own choice |
| How they are selected | by chance | by a price: forgoing the payment |
| What motivates them | they are paid for their time and given materials | they have a stake in the outcome |
| The force of the decision | advice to a parliament | the binding outcome of an election |
| The scale | a hundred people on one question | the whole electorate at every election |
| The route of adoption | at the parliament's will | a referendum |
5. What to take from it¶
Measured evidence for the main assumption. More than a hundred experiments show: when the few who have studied the matter decide, opinions shift towards the more considered, and the conclusion of a hundred people coincides with the referendum of a whole country. This is empirical support for the claim that narrowing the circle to the interested improves the decision rather than distorting it.
Selection by price instead of selection by lot. An assembly takes random people and makes them interested for a time — with materials, experts and pay. The protocol does the reverse: it keeps those who are already interested enough to have refused the money. The lot is fair, but its decision binds nobody; self-selection binds, because it is the election.
Deliberation Day shows what was missing. The idea of paying everyone for a civic act has existed since 2004 and has been adopted nowhere: it had neither a route of adoption nor a group that gains from it personally. The protocol has both — Routes A and B (045, 035) and a majority that wants the payment (048g §6b).
6. Weak point¶
The coincidence of the Irish assembly with the referendum is one case on one question; it does not follow that a small interested group always decides as the whole country does, still less that it decides better. The shift of opinions after deliberation was measured in people selected at random and given materials from both sides; those self-selected by stake are given no such materials. Correction of 01.10.2026: this does not make the transfer an assumption — a stake, as the accuracy-incentive experiments show (Bullock et al. 2015; Prior et al. 2015), cuts partisan bias by more than half, and the informed vote differently from the uninformed (Bartels 1996; Lau, Redlawsk); what remains is only that a "correct vote" in those studies is measured against the person's own preferences, not against the quality of policy. 🟡
Sources: Fishkin J., work on deliberative polling since 1988; Stanford Deliberative Democracy Lab; Citizens' Assembly (Ireland), report on the Eighth Amendment, 2017; results of the referendum of 25.05.2018; Convention citoyenne pour le climat, final report, June 2020, and the official implementation count; Ackerman B., Fishkin J., Deliberation Day, 2004.
Related: 019 §61.2–61.5 (the catalogue of alternatives, "a vision without a door") · 023 §23.2–23.3 (the summary table) · 036 §85 (ten reforms) · 013f (the formal statement) · 035 and 045 (the routes) · 048g §6b (who votes on the rules) · 006b (predecessors)