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.
23. Comparative Effectiveness — Single Digits vs Tens of Percent¶
Chapter: 05 — Empirical base and country cases Status: ⭐ The project's central persuasive argument (raised from the end of v6.55 — was §97.5) File version: v1 Date: 10 June 2026
23.1. Purpose of the section¶
Existing electoral reforms — RCV, open primaries, citizens' assemblies, technocratic instruments — yield an effect within single-digit percentages on each parameter. AB-EXIT yields an effect in tens of percent. That is a difference of an order of magnitude.
This section shows the difference quantitatively. Without it the comparison stays qualitative ("works / does not work"), whereas it should be quantitative ("works at 3–5 % vs works at 30–40 %"). This is the central persuasive argument for rational elites, for tech capital, for practising academics.
23.2. Empirical effects of existing mechanisms¶
Electronic voting (Estonia since 2005). Solvak & Vassil (Tartu): online voting raised overall turnout by 2–4 percentage points after ten years. Most online voters are people who voted offline anyway. Attraction of new voters — about 1–2 %.
Postal voting (Oregon since 2000). Berinsky 2001, Gerber et al. 2013: a turnout gain of 2–7 %. Noticeable but not structural.
Automatic voter registration (AVR). Brookings 2019: nine US states showed a turnout gain of 2–3 % among the previously unregistered.
Lowering the voting age to 16 (Austria since 2007). Zeglovits & Aichholzer 2014: adolescents vote at the same rate as adults. A gain of roughly 2–4 %.
Compulsory voting (Australia, Belgium). The strongest mechanism by effect — +25–35 % turnout. But it is not a solution to apathy and populism: a compelled electorate includes those who would not have come without a fine (about 23 % in Australia by the 2022 survey), plus 5–6 % of spoiled ballots. Correction of 30.09.2026: the earlier sentence "populist parties regularly win in Australia" is not borne out — One Nation received 6.4 % in 2025, and a party labelled populist has never entered government; by position (033c), though, both big parties work as populists, because everyone votes. Full turnout pulls the parties to the centre, but it is the centre of everyone including the compelled; it does not change the horizon of power (pensions, housing, the 2019 election) and creates no trust in parliament (28 % by the WVS). In detail — 023c.
Ranked Choice Voting (Maine since 2018, Alaska since 2022). Alaska 2022: the moderate Mary Peltola beat the populist Sarah Palin thanks to the redistribution of second choices. A local effect. The structural effect on populism — 0–3 % on average.
Open primaries (California Top-Two, Alaska, Louisiana). McGhee et al. 2014: they reduce the share of radical candidates in the final vote by roughly 5–8 %, but do not affect the composition of the electorate.
Citizens' assemblies (Ireland 2017, France 2020). An effect of a different order — an instrument for specific decisions, not an electoral reform. 0 % on electoral processes.
Democracy vouchers (Seattle since 2017). Brennan Center 2020: participation in funding widened from 1.5 % to 4.4 %. +3 % participation in funding, not in voting.
Education and quality press (historical). Putnam, "Bowling Alone", 2000: turnout growth of 5–7 % over 50 years. That effect has now reversed: turnout is falling, trust is falling.
Lobbying, corporate money, media machines. Bartels 2008, Gerber-Karlan-Bergan 2009: the effect of a large-scale media campaign on specific results — a 1–2 % shift at enormous cost.
23.3. Summary table of the effects of existing mechanisms¶
| Mechanism | Effect on turnout | Effect on quality | Effect on populism |
|---|---|---|---|
| Electronic voting | +2–4 % | 0 % | 0 % |
| Postal voting | +2–7 % | 0 % | 0 % |
| Automatic registration | +2–3 % | 0 % | 0 % |
| Lowering the voting age | +2–4 % | 0–2 % | 0 % |
| Compulsory voting | +25–35 % | 0 % (the centre of everyone, the horizon unchanged) | 0 %: populists neither grow nor fall |
| Ranked Choice Voting | 0 % | +2–5 % | +0–3 % locally |
| Open primaries | 0 % | +5–8 % | +3–5 % |
| Citizens' assemblies | 0 % | +5–10 % locally | 0 % |
| Democracy vouchers | +3 % locally | 0–2 % | 0 % |
| Education/press (historical) | +5–7 % over 50 years | +5–10 % | until 2000 |
| Lobbying/media | 0–1 % | 0 % | 0–2 % |
The strongest structural instruments are RCV (~3–5 % on populism) and open primaries (5–8 %). That is precisely the "single digit" to which existing reforms are confined.
23.4. The effect of AB-EXIT — another order of magnitude¶
A methodological caveat: AB-EXIT has been launched in no jurisdiction, so the figures below are model estimates, based on: - Analogy with the Alaska Permanent Fund Dividend (1976–present, 100 % participation) — see §14 - Behavioural economics (response to monetary incentives) - Mathematical models of the electoral process — see §13, game theory - Stress-testing with six AI models (449+ rounds)
These are justified estimates, not empirical facts. They require pilot testing in a real jurisdiction.
First effect — the structure of the electorate. In countries with turnout of about 50 %, AB-EXIT under a moderate model can take 20–35 % of voters out of the electoral process. The electorate changes by 20–35 percentage points — half of the remaining electorate.
At compensation of about $200–676 a year (1.5 % of median income), 30–40 % of current apathetics will accept option B. A structural cleansing of the electorate by 15–25 percentage points.
Second effect — the motivation of those who remain. If 25 % of voters exited via B, the remaining 75 % automatically have votes of weight 1.33 each. +33 % political influence for every remaining voter.
Third effect — a smaller base for populism. If 30–40 % of apathetics left via B, the populist mobilisation base has shrunk by 30–40 % of the original. An order of magnitude more than RCV (3–5 %).
Example: Brexit won with 52 % on a 72 % turnout. "Leave" was voted by 37 % of all voters. Most of them were mobilised apathetics. Had half of them taken £200–500 through AB-EXIT, the remaining 18–20 % for "Leave" would have lost to the EU's supporters.
Fourth effect — the planning horizon. The departure of the apathetic extends the electorate's average horizon from 7–9 years to 12–15 years — a twofold extension.
Fifth effect — the economics for honest politicians. §15 (PNAS 2020): corrupt politicians lose their structural advantage. An honest politician gains a competitive advantage of the order of +15–25 % in chances of winning.
23.5. Comparison by orders of magnitude¶
Existing mechanisms give results within single-digit percentages on each parameter.
AB-EXIT gives 15–40 % on every key parameter:
- +15–25 % structural change in composition
- +33 % increase in the weight of the remaining votes
- −30–40 % populist mobilisation base
- +50–80 % extension of the planning horizon
- +15–25 % competitive advantage for honest politicians
A difference of an order of magnitude. This is not "AB-EXIT is slightly better" — it is "AB-EXIT works in fundamentally different units of measurement".
23.6. Ready formulations¶
For communication with rational elites:
"You have instruments that give 2–5 % improvement. You fight populism tactically, winning a few percent in one state and losing a few in another. AB-EXIT, by model estimates, gives 20–40 % of structural effect. It is not an alternative to your instruments — it is another level of analysis. Your instruments fight symptoms; AB-EXIT changes the architecture. The cost of a pilot test is low. The cost of inaction is the next populist cycle."
For a tech audience this argument is especially strong, because they understand the difference between incrementalism (5 %) and structural redesign (50 %). They build their entire careers on the second.
One methodological point — the correct formulation:
"Existing mechanisms give empirically confirmed 2–8 % on various parameters. AB-EXIT by model estimates and analogy gives 15–40 %. This requires pilot testing in one state."
A tech audience values honesty about uncertainty. With this formulation the project looks like serious scientific mechanism design, not a utopian manifesto.
Related sections: - §13 — AB-EXIT game theory (the formal mathematical grounding of the 33 % Shapley effect) - §14 — Alaska PFD (the empirical basis of the 30–40 % estimate) - §11 — The paradigm from duty to choice (the emotional grounding of the same thesis)
Source: written in claude.ai chat, session 31 (7 June 2026) as §97.5 in v6.55. Raised to a prominent place in file 05 of the empirical base in v6.56 as the project's central persuasive argument.