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.
56f. Russia, September 2026: the Duma elections and what follows¶
Chapter: 09 — Other countries File: 09_056f · v1 · 22 September 2026 Source: post-election analyses of 21 September 2026 (Maxim Katz; Abbas Gallyamov; Meduza — Alexandra Prokopenko and Andrei Pertsev) and subsequent work with the architect
How to read this chapter (note of 02.10.2026). The text contains wording that is easy to misread: sums and formulas with coefficients are worked examples: the size of the payment is set as a percentage of the median income approved by referendum; a "reversible" choice means "at the next election": within one election exit is final; the efficiency bonus has been removed; a blockchain is one implementation option: the charter requires only open code. The exact-answers sheet 1d and the charter 048m are in force.
The State Duma elections of 20 September 2026 supplied material against which the repository's claims can be checked rather than merely derived. This chapter gathers the whole analysis in one place: what happened, what follows for the protocol, what was refuted, and which actions remain available.
The general case of the Russian opposition is in 056d; here only September 2026 and its consequences. The catalogue of falsification techniques tested against these elections is in 048i.5f, and the comparison of the authorities' behaviour before and after adoption in 048i.5g.
1. September 2026: the technique half-tested, and the test went badly¶
The State Duma elections of 20 September 2026 supplied the first material against which 056d.7b can be checked. The figures below follow Maxim Katz's analysis of 21 September 2026; the assistant did not verify precinct commission data, and the estimate of the "real result" belongs to the analysis, derived from "honest" precincts and exit polls. 🟡
| What | Figure |
|---|---|
| United Russia, party list | about 58 %, against an estimated real result in the ballot boxes of about 35 % |
| Single-mandate districts | 209 of 225; 100 % of districts went to candidates planned in advance |
| Total seats | 355 — an absolute record; the systemic opposition was left under a hundred |
| St Petersburg | exit polls gave Yabloko one mandate certainly and probably three or four; results were redrawn from scratch — United Russia around 20 % on honest precincts and a flat 55 % in the protocols |
| Moscow | protest turnout, queues up to an hour and a half after the precincts closed; the redrawing had to cover paper voting, not only electronic |
| Saratov district | 83 % for the Duma speaker; precincts with 99 % turnout and 98 % for him; the Yabloko candidate got 983 votes against his 208,000 |
The main point for 056d.7b, and it runs against us. The mechanism described in 056d.7b assumes a chain: the commission sees the true scale of dissent, common knowledge forms, friction appears. In Moscow and St Petersburg the first link fired by itself, with no sign at all: commissions saw a genuine protest result at their own precincts — and redrew it, in the analysis's phrase, "as if by a ruler," with no visible resistance. Seeing is not acting, and the scene of exchanged glances in 056d.7b must be stated more modestly than it is.
One distinction is necessary here, and it is a distinction rather than an excuse. What was overwritten was the count of votes — a number the commission produces itself and can therefore produce at any value. The sign in 056d.7b is built differently: nobody counts it, it can only be seen, and it cannot be "redrawn" — it can only be left out of the tally, while what was seen stays seen. September 2026 tested ordinary voting, not the sign. But it did test the weaker premise — that being seen by itself creates friction — and the answer was no.
What survives of the mechanism. Kuran was never describing immediate action: accumulated private knowledge changes behaviour not on the day but when an occasion appears, and half a year passed between the independent count in East Germany in May 1989 and the fall of the regime. That is true — and it is exactly why the technique should be described as measurement and accumulation rather than as an event. Said here once.
Three things in the analysis work for the chapter's theses, and all three are results rather than reasoning.
Novgorod region: a natural experiment nobody set up. It is the single region where Yabloko's list remained on the ballot for the regional duma — and it cleared 6 %. A poor, conservative region with below-average indicators and no liberal base. The reason, per the analysis, is mundane: the authorities were busy with the federal campaign and had no time to rewrite regional results. This is the best available measurement of what the mere absence of an alternative on the ballot is worth: where it was left in by oversight, it immediately took its percentage in a place where it "should not have existed." The chapter's thesis that visible support is manufactured gains here not an argument but a number. 🟡
Foreign precincts: a rout for the party of power. Where the administrative machine does not reach, the result inverts. For the protocol this confirms 048i.5b from the other side: an emigrated citizen remains a voter and acts when acting costs him nothing; in the protocol he "comes for the money," here he came to the polling station.
Boldness as the only measurable quantity. The analysis concludes that the result of an authoritarian election is predetermined, but the degree of brazenness in the fraud is the only thing that indicates the scale of what is intended. 355 seats and under a hundred for the systemic opposition is a bid for which a constitutional majority is not enough. The observation deserves recording on its own: a wholly drawn number still carries information — just not the information it claims. It says nothing about support and a great deal about plans.
A correction from Katz's second analysis (24 September): a drawn-in number binds whoever drew it. A result of 355 seats cannot be sold to the public and does not need to be: the constitutional majority would have survived with fifty fewer. It is addressed to one viewer. But a drawn-in number has a recipient who believes it — and thereafter it works as a mandate: if support for the war has "grown", then more can be demanded, and any argument about the war's price is met with the figure. The analysis calls this a debt pyramid: the more is added, the more must be covered. For the thesis of the examination this means the examination binds not only the examined but the examiner: a system that has written itself such a credit will be bent to service it.
And the strongest part concerns the thesis of elections as an examination. In 048i.6c the technique is described after Shulman: elections in an autocracy are the state machine's examination in loyalty, and digitising them abolishes the examination. The analysis gives this an operational form the repository lacked: "electoral mobilisation differs technically in no way from mobilisation to the front — it is exactly the same administrative exercise; the chain of command must on order turn people into loyal voters or into soldiers." And then the conclusion the examination is held for: mobilisation may follow the electoral map — the quieter the elections passed and the less plausible the drawn result, the better the local authorities are held to control their territory.
If so, the examination stops being a metaphor: its result is used as a target map. For 048i.6c this strengthens the argument about why the regime clings to paper elections and why digitising them is expensive for it — and it is also a warning about what an excellent grade on that examination brings to the inhabitants of the region that earned it.
Amendment after Katz's analysis of 25 September: rewriting after publication and the allocation of quotas. Tushino district 205: at ten precincts the list leader changed three days after the vote. Precinct 248 (Voikovskaya): posted and entered on 20 September — New People 224, A Just Russia sixth; on 23 September in the GAS system — New People 27, A Just Russia first with ~35 %. Precincts 410 and 415: exactly 100 votes each from United Russia to A Just Russia; precinct 405: 200 each from New People and from United Russia. The first wave drew too much for United Russia and pushed A Just Russia below the threshold, the LDPR sank below New People — the second wave took from New People and even from United Russia. The number in the protocol turned out to be not a result and not even a "drawn majority" but an allocation of licences among the sanctioned parties, adjusted after the fact. Two consequences for the protocol: a public copy does not prevent rewriting (048f.4: a copy is not a second entry), and forgery of the distribution inside A remains possible under the first-stage design — only tokens anchor it (048i §2).
Weak point. All figures come from a single analysis and are not verified against primary data; the estimate of "really about 35 %" derives from honest precincts and exit polls and is an estimate, not a measurement. Exit polls in authoritarian elections carry distortions of their own, and none of the numbers above can be reconstructed independently. The main limitation lies elsewhere: the analysis speaks to the diagnosis, not the remedy, and changes nothing in §8 — in today's Russia neither a referendum nor public discussion is possible, and the protocol remains an idea for after a transition rather than an instrument. This section cannot be read as an argument for the protocol; it serves as material on precisely what has stopped working. 🟡
2. The Meduza discussion: what the protocol solves and what it does not¶
The architect's question following Meduza's post-election discussion (Alexandra Prokopenko of the Carnegie Russia Eurasia Center in Berlin and Andrei Pertsev, 21 September 2026): how do the protocol and the voting methods proposed here address the problems named there. What follows is a point-by-point examination, including the points where there is no answer.
First, what is named as the problem. Per Pertsev the main instruments are not stuffing and carousels, which "do not give very much," but corporate mobilisation, controlled remote electronic voting, and the rewriting of protocols in territorial commissions. Mobilisation was tightened this cycle: brigade leaders were attached to HR staff, reporting was required, complaints appeared about demands to photograph the ballot and to vote remotely in a manager's presence; those who carry it out are paid bonuses. Pertsev argues directly against the formula "the ballots really are in the urns, so the elections are more or less honest": telling a man he will be fired if he does not turn up is not honest voting but a method of falsification, merely a more refined one. 🟡
| What is named as the problem | What the protocol does | The honest limit |
|---|---|---|
| Corporate mobilisation: turnout is produced by an employer under threat of dismissal | Strikes at the foundation of the technique: turnout stops being a commodity. Today a boss sells upward the turnout he produced; once everyone is paid for appearing, they appear without him, and the report "I delivered 95 %" is worth nothing (056d.5, 048h) | It does not abolish coercion over the choice, only devalues coercion over attendance. Coercion to take the sum is treated separately: it subtracts blindly (048i.5c) |
| Rewriting protocols in territorial commissions | Open votes are summed by token past the commission; an official figure below the sum of transmitted tokens is arithmetically proven fraud (048i.2) | Works only on the open channel and only with a high share of tokens transmitted |
| Controlled remote electronic voting | The token shows a person his own vote and is transmitted to a party; this is what remote voting lacks by construction | Full protection requires publishing all tokens; without it what remains is a lower bound, not accuracy |
| No valve: "whoever you vote for, it is forty shades of one party" | The protocol is not a party but a question for a referendum. People vote not for persons but for a rule of distribution, which requires no decent name to appear on the ballot | It requires a referendum to be possible. In today's Russia it is not (§8) |
| Fear: a man is afraid to vote on paper in case he is noticed | It changes the nature of the act: coming for one's own money is not a political deed and signals nothing. This is the same property the sign has (056d.7b): compatibility with coercion | The choice of recipient remains political; only the secret channel covers that |
| Rational conformism: "the rotten present is better" | It changes the arithmetic: today conformism buys safety and protest is risk without payment. The protocol puts money on the second scale, and puts it with those for whom it means most (§5) | It works on the apathetic and not on convinced supporters; it was never designed for the latter |
And one point deserves separating out, because only the protocol answers it. Prokopenko formulates the cause of conformism not as fear but as absence of a subject: "there is no clearly formulated alternative… nobody knows what a different future might be. So the rotten present is better than this unclear future of yours. And for protesting you can also get hit over the head."
This is precisely the empty slot described in §3 — but stated independently and more harshly: the absence of a formulated alternative makes conformism rational rather than cowardly. Against the fear of repression the protocol has no means. Against "an unclear future" there is exactly one means, and it is the protocol's principal one: a future expressed as a number a person can compute for himself. This does not make the protocol stronger than the police; it makes it the only item on the list that answers this argument at all. 🟡
And now what works against us. Pertsev closes with an observation: Putinism rested on a plebiscitary claim — one part of society was shown the other part supporting the course, and this was fixed at more or less genuine elections; that can no longer be said. Prokopenko describes the same construction as the signalling function of a constructed majority: "people saw each other at the polling stations, people know that they came, nobody knows how the neighbour voted" — and what results is the image of a majority supporting the course.
For the protocol this is bad news in two places. First, the signalling function is preference falsification (Kuran) in the form of a manufactured product, and the protocol destroys it only after adoption, having no effect at all before. Second, and more important: a regime that has stopped valuing appearances loses less by banning the protocol outright. While elections were made to look genuine, banning "transparent distribution of rent" was awkward (§7); once the pretence ended, the awkwardness went with it.
The conclusion on the architect's question, unsoftened: of the six problems named, the protocol is built so as to remove four and devalue a fifth, but not one of them is removed in Russia today, because every answer switches on after adoption, and adoption requires a referendum that does not exist. The Meduza discussion is a description of why §8 of this chapter is written as it is. The only thing that works before adoption and requires nobody's permission is the sign on the ballot (056d.7b), and its test in September 2026 went badly (§1). 🟡
3. The empty slot in the repertoire¶
The architect's conclusion from the September 2026 results: "so everything the entire Russian opposition came up with produced nothing at all. AB-EXIT would at least have been discussed in kitchens if the opposition had put it forward as its main proposal and spent millions, its viewers' resources and hours of video on it before the elections."
The architect's second remark sets the right frame: "this was in essence a report on how the opposition spent four years preparing for these elections — and here is the result." That is how it should be read: the analysis itself goes back five years and sums up a cycle rather than a single day. Seen that way it is not "a bad election" but the output of a completed preparation cycle: on the input side, millions of viewers, investigations, hours of video, accumulated infrastructure and a genuine protest turnout in two capitals; on the output side, zero of two hundred and twenty-five districts, not one unplanned mandate, the systemic opposition left under a hundred. For judging a repertoire this is the best possible setup: what is measured is not a single technique but the entire set, over a full term, with the resources that existed.
The first half needs qualifying rather than instant agreement. "Produced nothing" is true of today and false of the whole period: Smart Voting brought the opposition roughly twenty of the forty-five Moscow City Duma seats in 2019, and that was a result, not a gesture. The accurate statement is different: the techniques worked right up to the point where closing them became worthwhile, and by 2021 they were closed — the app was removed from the stores and the results were rewritten.
But the whole repertoire shares a structure, and it explains why the collapse was simultaneous. Smart Voting, observation, exit polls, spoiling the ballot, boycott, "Noon against Putin" — these are all ways of expressing a preference inside a count the authorities themselves produce. They differ in tactics and coincide in their point of failure: each one ends at a number the commission writes. While the commission was still embarrassed, the techniques meant something; in September 2026 the embarrassment ended, and they all went to zero at once rather than one by one. This is not a failure of ingenuity but a property of the class.
And this is exactly why the single exception looks so convincing. In Novgorod region Yabloko took over six per cent not because a tactic was found but because the list stayed on the ballot and there was no time to rewrite it (§1). What worked was not a voter's action but the presence of an alternative. That points to where the repertoire is empty: not in techniques of voting but in what there is to vote for.
Now to the supposition itself, and the answer splits in two.
As an instruction to a voter in the booth — no, and this must be said plainly. Smart Voting came with an instruction executable in the booth. The protocol has no such instruction: it requires adoption, by referendum or by law, and neither is available today (§8). An opposition that made the protocol its main proposal before the elections could not have answered the question "what do I mark on the ballot on the twentieth of September".
But the test "an executable instruction" is set for one addressee — the voter — while an election has a second one, whose decision is built differently. That is examined in §3c, and it cancels part of the conclusion drawn here.
As a subject of conversation — yes, and here the architect's position is strong. By the analysis's own diagnosis, the failure lay not in how people voted but in the fact that they did not come: "the country passed the test of passive loyalty; it makes no difference why people do not resist." Apathy is not cured by improved booth tactics — tactics answer "how," and the question left unasked was "why bother at all." And here the empty slot becomes visible: everything the opposition offered was procedural or moral — honest elections, against the war, against corruption. An offer of the form "here is your share, here is what you personally get" has never been made by anyone — and the slot stood empty for the whole cycle, with the money and the audience available. This is §6 of this chapter seen from the other side: not "why won't the opposition take up the protocol" but "what occupies that slot instead" — and the answer is nothing.
And the supposition has a testable version, closer to hand than it looks. The protocol requires adoption; the sign in 056d.7b does not. The sign is the one item in the whole construction that could have been announced the day before and performed in the booth without a single change in law. It was not announced; the architect himself regretted this on seeing the election date. So the supposition that the opposition could have spent resources on this and got something is untested rather than refuted, and those are different things. A testable version exists and is cheap — hours of video and a single day.
Weak point. "Would have been discussed in kitchens" is not a result by this repository's own standards, which elsewhere demand a number. Conversation does not rewrite a precinct protocol or change the composition of the Duma; it changes what people know about one another (056d.7b, East Germany 1989), which works slowly and without guarantees. The relationship "size of payment → participation" has not been measured for Russia, but it has been measured where the payment exists (§13, first row); what has been measured nowhere is the effect of a payment tied to the act of voting. 🟡
3c. The second addressee: the schoolteacher at the commission table¶
The architect's objection to §3 and §3b:
"The protocol would have increased the doubts of state employees and given them cause to think. The schoolteacher on the commission needs money, she is for the authorities — and all they give her is threats and handouts, and they make her work at the elections. And here she can honestly take money, and even the opposition is in favour of giving money to her rather than to them. And all her colleagues are for it, and the parents are for it, and her husband is for it, and the neighbours are FOR it. And who is against, and why is it illegal or impossible to adopt? Does the opposition offer anything like this now?"
The objection strikes at the construction of the test rather than at its result, and so cannot be dismissed with an adjustment for scale.
The error was in the choice of addressee¶
"An executable instruction" is the right test for a voter: he has one action, it is performed in the booth, and for him in September 2026 the protocol indeed offers nothing. But an election has a second person, and her decision is not a ballot. A member of a precinct commission decides not "for whom" but "whether to sign what has been written".
That decision requires no instruction and is not prepared for in advance. It is taken at a table, at the end of a long day, looking at the people beside her. Everything that acts on it acts before that moment, and through how she understands her own position.
The composition of this electorate in figures — by sex, age and demography — is gathered in 019e.7 (Russian only for now); here is one person from it.
Her position, without sympathy and without reproach¶
The schoolteacher at the commission table is neither a supporter nor an opponent. She is a state employee, she was assigned to the election, the work pays little, the result is demanded of her, and refusing the assignment costs more than accepting it. Shulman's loyalist majority (056c) is built exactly this way: loyalty consists not in acting for the authorities but in not participating against them.
And this is what the state gives her today: threats and occasional handouts. Neither is a right, both depend on a superior, and both must be earned. That is a precise description of a relationship in which there is no share — only favour (§9: a payment without a formula is a handout, not a right).
What the entire current repertoire offers her — and it is zero¶
| The opposition's technique | What it is to her |
|---|---|
| Observation | she is the adversary; she is watched, she is caught |
| Exit polls, "honest precincts" | she is the object of measurement; her work is the thing being exposed |
| Smart Voting | addressed to her as a voter, not as the person at the table; it does not touch her decision |
| Boycott, spoiling ballots | nothing |
| Corruption investigations | about somebody else's wealth; they change nothing in her position and promise her no share |
The whole repertoire addresses her either as an enemy or not at all. This is not peculiar to her case: the general form is in 040i.2b, which shows why the repertoire is by construction addressed to those who can afford it. Not one technique offers her a gain — only risk if she deviates and exposure if she does not. That is the answer to the architect's question: nobody offers anything of the kind at present.
What the protocol offers — three things, and none of them an instruction¶
First: a number about herself. Not "the authorities steal billions" but "your family's share this cycle is such-and-such". The difference lies not in volume but in the addressee: the first is about someone else's, the second about her own. The calculator (§5) is exactly the translation of an accusation into an account.
Second: a question her superiors cannot answer aloud. "Who is against my being paid my share, and why can it not be adopted?" — the foreman test (23b.2) put not to a fitter but to a head teacher. There are still two answers, both losing: "the state does not want to hand it over" or "asking is disloyal".
Third, and the architect puts this most strongly: coincidence instead of isolation. Normally a position against the authorities leaves a person alone within her own circle — the spiral of silence rests on that (056c). Here the circle coincides: colleagues, pupils' parents, husband and neighbours want the same thing, and wanting it requires neither courage nor a political position. For the first time, being in favour of what the opposition proposes does not remove a person from her milieu but places her in the middle of it.
And this explains why September failed in the way it did¶
§1 recorded against us that the commissions saw the genuine result and rewrote it without friction — "seeing is not acting". The conclusion was right and the cause was not named.
The cause is what exactly they saw. They saw a fact about strangers: so many unknown people voted otherwise. Such a fact creates no friction, because rewriting takes nothing away from whoever is looking.
Friction is created not by information about others but by one's own claim, destroyed by that same act. The protocol gives her precisely that: in signing a rewritten protocol she signs against a sum owed to her family. This does not turn her into a hero — it turns the signature from a dull procedure into a personal loss.
What it does not give — and the boundary must be held firmly¶
The likely outcome is not refusal but a signature with a grievance. She will sign: refusal costs her job, and the protocol will be rewritten without her. September showed that a grievance is not enough for action, and 048f.4b records the same against us.
But by the standard of comparison with the status quo (048f.4b, 040b.6) a grievance is a result rather than a zero, because the alternative to it is not "refusal" but the present state, in which she is offered nothing at all. And that is the only honest way to measure: not "will she break the machine on the twentieth of September" but "does anything change in the position of the drive belt's last link". It does.
And here is the price she is held at¶
The architect's refinement, which moves this section from reasoning into arithmetic:
"Important: the authorities do not maintain all their supporters on large money. In the villages and the regions it is negligible money and fear."
The figure is checkable, and it turned out lower than one might have supposed. At the State Duma elections of autumn 2026 the rates for precinct commission members are: chairman — 63 roubles an hour, deputy and secretary — 57, an ordinary member — 45 roubles an hour [verified against September 2026 publications; the assistant did not open the CEC resolution setting the rates].
Forty-five roubles an hour is below the statutory minimum wage, and the lawfulness of that had to be explained separately by the Supreme Court [second-hand: from a publication's headline; the assistant did not read the ruling].
Compute what can be computed. Three-day voting with preparation and an overnight count comes to something like forty hours of work. Forty hours at forty-five roubles is about 1,800 roubles for the whole election.
| What | How much |
|---|---|
| An ordinary commission member's earnings for the entire campaign | ~1,800 ₽ |
| The dividend per person per cycle (a repository estimate, unverified) | ~11,000 ₽ |
| For a family of three | ~33,000 ₽ |
| Ratio | roughly twentyfold |
Three things follow from that figure¶
First: loyalty here is not bought. Forty-five roubles an hour buys nothing — it merely formalises participation. What holds her is not the pay but the fact that refusing costs more than agreeing: her job, her relations with the head teacher, a reputation for being difficult. This is precisely Shulman's definition of the loyalist majority (056c) expressed in money: loyalty consists not in acting for the authorities but in not participating against them being cheaper.
Second: the regime can outbid — but not the people it needs to. The first edition said here that the regime cannot outbid; that is arithmetically wrong, and the error must be corrected plainly.
The size of a precinct commission is set by statute according to the number of voters: up to 1,001 voters — three to nine members; above two thousand — seven to sixteen [verified against the text of the law]. Precincts number in the tens of thousands and are decreasing [second-hand]. Even on the upper estimate this concerns hundreds of thousands of people, not tens of millions.
Raising their remuneration from one and a half or two thousand roubles to thirty thousand comes to something of the order of ten to thirty billion roubles a cycle, a small fraction of what the repository estimates the dividend itself to cost (§5). The budget has that money.
And precisely for that reason the move does not save anything. What can be bought is the tellers, not the queue. The drive belt is not three hundred thousand people at tables but tens of millions of administratively dependent voters who have to be brought in; buying their loyalty by the same logic means paying everyone, and that is the dividend. The outcome of the move: a commission honestly and accurately counting an empty room.
Third: and fear does not need outbidding — it needs bypassing. The protocol asks her for no courage and does not ask her to refuse anything. It gives her a lawful ground, issued by the state itself (056, irreversibility 4). What works against fear is not greater bravery but a piece of paper that removes the need to show any.
The staffing constraint: where there is nobody left to dismiss¶
The architect's correction continues, and it changes the geography of the argument:
"And where you cannot dismiss the last teacher? Somewhere there will be refusal, and they will go looking for half-blind grandmothers — who will agree, but physically cannot stuff ballots. And the grievance will be an angry one, and so will the sabotage."
A threat of dismissal works only while a replacement exists. That condition is being met ever less well in the Russian countryside: as of April 2025, 36 % of all open teaching vacancies were in rural schools (6,574 of 18,287), the shortage traditionally concentrates in rural districts, and teacher numbers have fallen for five years running [verified against publications; the assistant did not open the ministry's primary statistics].
Hence an asymmetry the repository did not have:
Coercion is weakest exactly where the demand for compliance is highest.
It is the rural and small precincts that deliver the results for which the machine exists: §1 cites the Saratov district with precincts showing 99 % turnout and 98 % for a single candidate. And these are the same territories where a head teacher cannot dismiss the only subject teacher, because there is nobody to replace her with and the district would be left without physics.
Replacement degrades the product, and this is mechanics rather than morals¶
If the teacher does refuse, the commission has to be staffed by somebody. The statute sets the floor: a small precinct has only three to nine members [verified], so the loss of one or two is already a noticeable share of the body.
And here something usually left unsaid must be named: falsification is labour. Rewriting a protocol "as if by a ruler" (§1), reconciling the figures, meeting the deadline, producing no discrepancies — that takes literate, quick and organised people. The teacher is exactly that; it is why she is used.
Replacing her with compliant but unfit labour — a pensioner who will agree but cannot keep up and cannot manage — degrades the quality of the forgery, not merely its speed. What the regime loses is not loyalty (it never had that) but skilled coercible labour, and there is nothing to replace it with: a skilled volunteer for this work is rare, and there is nobody left to compel.
Katz's second analysis: the director is squeezed from both sides, and he pays first¶
The architect's remark on the second post-election analysis (Maxim Katz, 24 September): "Katz again confirms the dependent electorate, and confirms that the protocol can effectively drive a wedge between the regime's commands and its accountable, dependent executors."
The first — yes, and with details the chapter did not have. The second — no, and this must be recorded precisely: the analysis confirms the location of the wedge and at the same time confirms that in September 2026 there was no wedge.
What the analysis adds to the portrait of the dependent. Three things, all observed.
Tracking by name. "This time the managers did not merely persuade en masse but took the trouble to check up on and track each person individually — voted or not." Individual monitoring of turnout is not a future risk of the protocol but the present state of affairs. Public A/B status (048h.2) does not create surveillance that does not exist; it changes what stands behind the tracked act — a lawful ground rather than a fault.
The family multiplier. "Social media are full of stories of a state-employed mother tearfully begging the whole family to vote, or she will face trouble at work." One dependent person yields several compelled votes: coercion passes through the household. This is 019e seen from the other side — a guarantee tied to one person drags along everyone who depends on her. Under the protocol the same channel runs the other way: a mother with a lawful ground not to come has no need to beg anyone, and the multiplier changes sign.
A vote so as to be "left alone". "These are not votes for the war, they are votes so that they and their relatives are left alone." The analysis independently confirms what this section derived: the dependent person has no position — he has a price of non-participation, and it is paid with a tick.
What the analysis adds to the portrait of the executor — and this is a new link. "Hundreds of thousands of petty bosses across the country are not landlords in an ivory castle. They are people on low pay with enormous responsibility, who depend entirely on their collectives. A head teacher who pressed, shouted and threatened colleagues still has to work with those people. A collective that has turned against its director will make his work unbearable."
Until now the foreman in this chapter (23b.2) stood as the transmission link between the vertical and the subordinate. The analysis shows the link is squeezed from both sides: a demand from above, and from below a collective on which the director depends after the election just as the collective depends on him before it. And he risks his relations with the collective "in the name of the state's will".
Hence the exact location of the wedge. It is not between the director and the state — there the director is defenceless, and that is known. It is between the director and his own people, and the director pays for it: a collective that has acquired a question the director cannot answer (23b.2) makes his work unbearable — that is, it exacts from him the very price the analysis describes as an already existing risk.
And why "confirms that the protocol will drive a wedge" is wrong. The analysis ends: "The state vertical is loyal and works like ChatGPT. Whatever demands were loaded into it, those are the figures it produced. No failures, no sabotage on the ground, no improvisation." That is a direct observation: the director was squeezed from both sides — and complied all the same, because nobody was pressing him from below. There is somewhere to drive the wedge; there was nothing to drive it with.
What was missing was not pressure but a ground for pressure that the collective can name aloud without becoming the opposition. The protocol supplies exactly that: "I am owed this, and you are demanding that I give it up." This is not confirmed by the analysis — it is what follows from the analysis once the thing absent in September is substituted in.
On sabotage: weaker than stated, and the more reliable for it¶
"Angry sabotage" is a claim about behaviour nobody has observed, and September 2026 is direct evidence against it: the rewriting happened without friction (§1). It cannot be recorded as a forecast.
But what was stated is not needed. A withdrawal of discretionary effort suffices. A forgery rests not on instructions being followed but on people doing more than required: staying late, finishing the count, covering, finding solutions in the last hour. Ceasing to do more than required is not sabotage but working to rule; it cannot be punished, because from the outside it is indistinguishable from conscientiousness.
That is the plausible mechanism, and it requires neither courage nor anger — only the absence of a reason to try. A grievance is precisely what removes the reason.
Katz's second analysis confirms this outright: "no failures, no sabotage on the ground, no improvisation" — even though the directors, by his own account, were risking their collectives. The link squeezed from both sides complied in full, because there was nothing from below worth not complying for. That is an argument for the weakened formulation above and against "angry sabotage": without a ground from below there will be no working to rule either, and with one, working to rule will suffice.
And the last link: "the budget grows and there will be nothing left to steal"¶
Here the true must be separated from the overstated.
It is not true that nothing would be left to steal. Ten to thirty billion a cycle in commission remuneration is a quantity invisible in a federal budget. It does not consume the rent, and to claim otherwise would be the very stretch that 23b.1 was written against.
What is true matters more. Payment for loyalty, converted from fear into money, acquires three properties fear does not have:
| Fear | Money | |
|---|---|---|
| visibility | recorded nowhere | a budget line with a name |
| reversibility | can be relaxed for free | by irreversibility 1, cannot be withdrawn |
| what it admits | nothing | that turnout is paid for, and therefore not voluntary |
The third row is exactly the "it will become more visible that the authorities pay for loyalty" that the correction points to. This is not the ruin of a budget — it is a public price list for obedience, which until now has never had to be named anywhere.
Hence the exact formulation of the asymmetry¶
The state pays her forty-five roubles an hour to produce a number. The protocol offers her family a sum of the order of twenty such earnings — as a right, not as a fee, and with the agreement of everyone around her.
This is not a competition between two offers. They are different categories: one is payment for a service that cannot be declined, the other a share that need not be earned. And the regime cannot enter that competition without ceasing to be itself.
A caveat about the regions, and it favours the architect. The maximum compensation for time away from one's main job is set by region and in Moscow in 2026 reaches 180,861 roubles for a full month [verified against publications] — but that is not remuneration, it is reimbursement of forgone salary, and it is computed from average earnings. In a village where average earnings are a fraction of Moscow's, the reimbursement is a fraction too. So the poorer the territory, the cheaper its drive belt is to the regime — and the weightier, for that same person, a fixed dividend computed from the national median rather than from her own wage.
Weak point. All of the above reasons about the motives of a person whose motives nobody has measured; the portrait was assembled by the assistant from the repository and from the general description of the loyalist majority, not from data about precinct commission members, which he did not seek. The architect's claim that "all her colleagues are for it, her husband is for it, the neighbours are for it" is an assumption about support for a payment nobody has seen; the nearest confirmed thing available is Alaska's take-up of about 85 % of residents (014), and it says that people take money, not that they speak about it openly where it is promised the label of extremism. And most vulnerable of all: the step from "grievance" to "friction" is shown neither here nor anywhere else, and September 2026 is a direct observation against it. Separately on the commission rates: forty hours of work is the assistant's estimate rather than a norm, and a member's actual hours vary by region and by the number of voting days; the twentyfold ratio rests on a dividend figure flagged as unverified in §5, and changes proportionally with it. 🟡
3b. What the whole opposition cost the authorities — and where the difference lies¶
The architect's remark after all the counter-moves had been examined:
"Put in the repository that under any scenario everything gets much harder for the authorities. At these elections — how did the opposition hinder them? Not at all."
The second half is exact and is borne out by the material of §1 and §3. The first must be recorded more weakly than it was stated, because a counter-example exists and we wrote it ourselves (048f.4b).
The baseline: what September 2026 cost the authorities¶
| What the opposition did | What it cost the authorities |
|---|---|
| Observing at precincts | nothing: with observers present the results were rewritten "as if by a ruler" (§1) |
| Exit polls and "honest precincts" | a clash of estimates against an incontestable piece of paper; it never left its own audience |
| Smart Voting | it worked until 2021, then the app was removed from the stores and the results rewritten (§3) |
| Calls not to turn out | a negative price: the abstention of the disloyal is the aim of the machine's second operation (056, correction of 2026) |
| Spoiling ballots, "Noon against Putin" | nothing in the precinct protocol |
| Four years of preparation | 355 mandates, 209 districts of 225, fewer than a hundred for the systemic opposition (§1) |
The cause is common to them all and is named in §3: every one of these techniques expresses a preference inside a count produced by the authorities themselves. Differing in tactics, they coincide at the point of failure — each runs into a number the commission writes. While the commission was still embarrassed they meant something; once it stopped, they went to zero together.
What the authorities pay on each branch of the protocol¶
| The authorities' move | What they pay | Where analysed |
|---|---|---|
| Stay silent, do not adopt | the instrument stays on the table for any transition player: no programme or coalition needs inventing | 056b.3 |
| Ban it | the first public ban on handing money to the people; every word of the ban works as agitation | 056b.3 |
| Adopt it universally | the five irreversibilities; both passenger flows are dried, not one | 056, correction of 2026 |
| Adopt it while excluding the dependent | there is no working form: an equivalent payment is taken and the vote skipped, a smaller one produces a grievance announced by the foreman | 056, correction of 2026 |
| Adopt it and order people to take the money and turn up anyway | a mass accounting discrepancy, which is budget embezzlement with traces held by banks | 048f.4c |
| Adopt it and use the register of exits as a target map | nothing: this is the one move with a gain | 048f.4b |
Hence the exact formulation, and it is not "under any scenario"¶
The last row is a counter-example, and we wrote it ourselves as the falsifying condition for the loosening claim. So the honest version reads differently:
No free moves remain to the authorities. Exactly one winning move survives, and it is bought at the price of maintaining the very register that compromises the count.
This is zugzwang in the exact sense of 001c: one must move, moves exist, each costs something. Not "everything gets harder" but "nothing comes free any more".
Where the difference actually lies — and this is the point of the section¶
The comparison should be of the point of failure, not of outcomes.
| The opposition's repertoire | The protocol | |
|---|---|---|
| where the technique ends | in a number the commission writes | in the budget, the payroll, a bank entry, the foreman's mouth |
| who has the last word | the very party the technique is aimed at | the treasury, a bank, the statute, the neighbour on the landing |
| what must be broken for the technique to fail | rewrite the protocol | break the books, or say aloud what cannot be said |
The whole value lies in one thing: the opposition's techniques end where the authorities hold a monopoly, while moves against the protocol end where they hold none. That is the answer to "how does the protocol differ from Smart Voting", and it is not about effectiveness but about jurisdiction.
A caveat without which the tables compare incomparable things. The left column is what happened; the right is what would happen after adoption, and adoption is unavailable in today's Russia (§8). This is a comparison of construction, not of outcomes, and it must not be read as "the protocol would have done better than Smart Voting on 20 September": on 20 September the protocol would have supplied no executable instruction at all (§3).
Weak point. The "what they pay" column consists of derived prices, not measured ones: no branch has been observed anywhere, and the magnitude of each price is unknown — "costlier" in the table does not mean "costly enough to change behaviour". September 2026 showed that a price can be paid and change nothing: the rewriting happened in front of observers and carried no consequences. And the summary itself was compiled by the assistant from his own sections, that is, it checks the repository against its own material — not one row has external verification. 🟡
4. What could have been done over the cycle: a plan drawn from the three analyses¶
§3 established that the slot is empty and that the architect's supposition is untested rather than refuted. This section sets out what exactly "millions, viewers' resources and hours of video" turn into if directed at the protocol. The order follows what the analyses themselves named as the problem, not what suits us.
A caveat without which the plan becomes a lie. Nothing below wins an election under the present regime or cancels §8. Everything listed is work for a transition window and for measurement, not for polling day.
| No. | What to do | Against what precisely | Cost |
|---|---|---|---|
| 1 | A share calculator (specification in §5). A person enters his region and household and sees his monthly sum. Not a text, not a programme, but a number about himself | Directly against Prokopenko's central point: "there is no clearly formulated alternative; nobody knows what a different future might be" (§2). An abstraction does not compete with "the rotten present"; a number does | weeks of work, no permission required |
| 2 | The measurement nobody has. A survey: do you know how much rent per person there is; would you support a rule distributing it; would you vote for whoever proposed it | Against §3's own weak point: the relationship "size of payment → participation" is measured in Alaska (§13) but unmeasured for Russia, and a payment tied to voting has been tried by nobody | the cost of a panel survey |
| 3 | A ready referendum question, legally drafted, with the share's unconditionality and the formula's protection (004.4.3c, 004.4.7b) | Against the prospect that in a transition window discussion starts from zero and consumes years the window does not have | lawyers' work, once |
| 4 | The ballot sign as measurement (056d.7b) | The one item in the whole construction that works before adoption and requires nobody's permission. It was available in September 2026 and went unused | hours of video and a single day |
| 5 | The main line of defence: the rule must have no owner. Spread through those who cannot be struck off — The protocol is a rule, not a person: anyone may voice it, including regional candidates of systemic parties who already ran slogans "for peace" and "for a free internet" (§2) | Against layer 1 (048i.5f): a candidate is struck off, a rule cannot be — it has no surname | zero |
| 5b | The repository as short clips, one argument per clip. The architect's remark: "everything in the repo can be made into small videos, and each person watches what appeals to him; there is material there that even pensioners will find interesting" | Against the fact that one long speech cannot address everyone at once (§3): fifty short ones can, and the viewer selects. This also answers the "party of the big cities" objection: the pensioner is addressed by late payment and the arithmetic of "two pensions," the small trader by two or three per cent of turnover, the state employee by turnout ceasing to be a commodity | editing, no permission required |
| 6 | Make nobody a loser. The share goes to everyone without exception, officials and war participants included | Against the lesson Gallyamov draws about himself: he declines to push recovery from war participants because it would create a group ready to defend the regime to the death (056d.5b) | discipline in wording |
What of this was available in September 2026 and not done. Points 1, 2, 4 and 5, entirely. None required registration, accreditation, or presence in the country.
The "loud or quiet" fork turned out to be false, and the architect removed it. The assistant had put it this way: loud advocacy delivers mass but invites a ban; quiet seeding preserves the subject but builds no mass. The objection: "banning it is very hard, and the attempt itself is noise and publicity. And after a very illogical and glaring prohibition the quiet seeding gets closed just the same — only without the popularity."
Right on both halves, and the assistant erred precisely where §7 of this chapter gives warning.
First: a ban here is expensive, and expensive by the nature of the subject. To ban a rule you must name it. And named, it sounds like "distributing subsoil income among citizens is extremism" — while Article 9 of the Constitution calls the subsoil "the foundation of the life and activity of the peoples" (§7). The comparison with Smart Voting shows the difference: there what was banned was an app, a domain and a list of surnames, all of which have an address and can be removed by an administrator in silence. A rule has no app, no domain and no surname; banning it has to be done out loud.
Second: quiet seeding is banned more cheaply, not less often. The assistant treated silence as protection. It is not: the mechanism of the quiet targeted ban is described in 019c — benefit concentrated, cost spread thin, the apathetic voter noticing nothing. Such a ban works best precisely against whoever has no supporters: no noise, no cost, an administrative footnote. So silence does not avert the ban, it cheapens it.
Hence the correct formulation, the reverse of the earlier one: loudness does not invite the ban, it makes it expensive; silence does not protect from the ban, it makes it free.
And a third thing the fork lacked entirely: what is dangerous is not loudness but ownership. Smart Voting was killed not by the app being blocked but by the label that attached to the organisation — and through it to every associated idea. Hence the construction's real defence: the rule must have no owner. A rule voiced by economists, by regional deputies of systemic parties, by local activists and by anyone else cannot be killed by killing any one of them; "those people's plan" dies with them. This is exactly point 5 of the plan above, and it is now clear that it is not a technique of distribution but the main line of defence.
And finally, what turns the ban from a risk into an outcome. §7 of this chapter already states that to ban a direct dividend the authorities would have to admit that the Constitution is a fairy tale for plebeians, and that admission is itself the confirmation. So the loud branch wins either way: either the subject spreads, or it is banned — and the ban says something about the subject that no amount of campaigning could say. The quiet branch has neither.
Weak point. The plan rests on the assumption that the empty slot is fillable by this protocol in particular; point 2 exists precisely to test that assumption rather than confirm it. The Alaskan measurement (§13) says money moves the young and the less well-off in particular — but says it of a democracy where the payment is not tied to voting, and so cannot stand in for a Russian measurement. And point 2 runs into the rule adopted in §10: a survey about money frightens less than a survey about politics, but remains a survey — its result serves as a lower bound, not as a measurement. 🟡
5. The share calculator: a specification¶
The architect took the plan's first point and turned it into an object: "in a video you could say — download an Android calculator of payments by city or by country and by household, and see a monthly salary or two pensions in it."
This is materially stronger than what was written, for three reasons, and one of them changes the plan itself.
First: it moves the argument from speech into a thing. A video is watched once; an app stays on the screen and is opened a second time when petrol goes up. A clip has to be rewatched; a number has to be recalculated, and the person recalculates it himself.
Second, and this is the main find: the unit is not roubles but "a monthly salary" and "two pensions." Abstract thousands mean nothing; a familiar quantity means everything. And from this follows a breakdown by city that the plan lacked: rent per person is computed nationally, but its meaning is local. The same sum is small change in Moscow and a salary in a district town. An app that shows the share beside the local median wage tells different people different things — and speaks loudest to exactly those the construction is aimed at (§5: for a nurse in a district hospital the sum changes everything, for a Moscow hipster it does not).
Third: point 1 of the plan swallows point 2. The measurement "nobody has" comes out of the same app: where it is installed, what households are entered, whether people return. No separate survey budget is required — but this carries a cost, discussed below.
What has to be inside¶
| Requirement | Why |
|---|---|
| The formula is open, not only the result | A calculator that emits a single number is propaganda and will die on first contact with an economist. The computation must be visible: budget income from rent, what is deducted, how many people it is divided among |
| The figures can be changed | A sceptic who enters his own percentages and sees his own result stops being a sceptic and becomes a user. The objection "you have inflated everything" is answered before it is raised |
| A source for every line | The data come from the published budget; a link beside each number |
| Computes on the device, with no server | See the security note below |
| Open source and open data | Otherwise "this is a fake by foreign agents" is a free answer |
| No account, no registration, no donate button | The same |
Two honest limits¶
Distribution, with the architect's correction. The assistant recorded the app store as the weak point. The objection: "Apple's store will not remove it quickly, or will refuse outright, and Google too — on what grounds? All the app contains is figures."
The correction is accepted, and it follows from the same logic as §4. Removing Smart Voting was easy: it was a list of surnames presented as voting guidance, and a takedown demand for it fits in one line. Here what is proposed for removal is a calculator that divides published budget figures by the population. It contains no call to action, no surname, no organisation; the demand would read "remove the app that divides published numbers," and stores carry hundreds of such apps. Refusing on that ground is awkward, and complying is awkward too: the precedent is visible to everyone.
So a store should be treated not as the weak point but as a tripwire: if the app is removed after all, that is itself news, and the news is that a calculator was removed. For the cause that is no worse than installation — the same mechanism §4 describes for bans in general.
Both fallback forms remain, not out of pessimism but because they cover different failures: a web page that works offline after first opening (nowhere to remove it from, installs by link) and an Android package (survives the site being blocked). The store gives reach, the page gives indestructibility, the file survives blocking; together they insure rather than replace one another.
Security against measurement, and this is a genuine conflict rather than a caveat. An app that asks a Russian resident about political preferences and sends the answer to a server is dangerous to him. So the computation happens on the device and sends nothing; any measurement is voluntary, anonymous, aggregate-only and announced on screen before the tap. That is markedly worse than a proper survey. The choice is between data quality and the user's safety, and here the choice is obvious; point 2 of the plan remains a separate task for panel surveys outside the country, and the app yields only a coarse picture. 🟡
The second scale: what to compare against¶
The architect's proposal: "you could compare it with what the budget spends on maintaining deputies, or any other line on bureaucrats — people will be surprised, it is a very small sum."
The figures were checked, and they yield a stronger conclusion than expected, because it runs the opposite way to the customary one. The denominator is Russia's resident population on 1 January 2025, 146,119,928 [Rosstat, verified].
| Line | ₽ per person per year | ₽ per month | Source |
|---|---|---|---|
| Oil and gas budget revenue, 2024 — a flow past the citizen, not a payment (see below) | 76,177 | 6,348 | Minfin, 11,131 bn ₽ [verified] |
| Oil and gas revenue, 2025 | 58,014 | 4,834 | Minfin, 8,477 bn ₽ [verified] |
| Section "General state matters," 2025 | 15,404 | 1,284 | Law 419-FZ, 2,250.9 bn ₽ [verified] |
| The entire Federal Assembly, 2025 | 154 | 13 | 419-FZ, 22.56 bn ₽ [verified] |
| — of which the State Duma | 102 | 8.5 | 14.84 bn ₽ [verified] |
| — the Federation Council | 53 | 4.4 | 7.72 bn ₽ [verified] |
| The "palace" from the 2021 investigation, one-off | 684 | — | the claimed 100 bn ₽ [a claim, not an audit] |
| Damage from corruption offences, 2024 | 208 | 17 | 30.4 bn ₽ [second-hand] |
What follows, and it is inconvenient for the customary agenda. Maintaining the entire parliament costs a citizen thirteen roubles a month. The palace around which the decade's best-known investigation was built is six hundred and eighty-four roubles once, about one per cent of one person's annual share. And the rent is six thousand roubles a month, every month.
The ratio: oil and gas revenue per person is 375 times the cost of both chambers of parliament.
So for fifteen years people were shown numbers that divide into nothing. Outrage per rouble in them is enormous; roubles per person are not. This is the arithmetic version of the same empty slot as §3: the opposition's material was moral ("they steal, and it is vile") rather than countable ("here is your share"). And it struck a target not worth striking: taking everything from the deputies leaves a person thirteen roubles a month.
An honest correction without which the conclusion becomes a lie. "Bureaucrats cost pennies" is true of parliament and false of the apparatus: the section "General state matters" is 15,404 ₽ per person per year, about a quarter of the rent. That is real money. The precise formulation: the visible, emotionally charged targets are negligible, while the apparatus as a whole is substantial — and still four times smaller than the thing nobody talks about.
Hence the calculator's second screen: beside the share, show not only the local wage but what people are customarily invited to be angry about. The first screen answers "how much is this for me," the second answers "why was I shown the wrong number for fifteen years."
Weak point. Section 01 spending is the planned figure from the budget law: actual execution by section could not be found, and the law's appendices contain only the open part. The palace figure is an investigation's claim, not an audited quantity. The corruption damage figure could not be confirmed at source. And above all: the comparison is sound arithmetically and does not make theft any less theft — it says only which number is worth showing if the aim is to move the apathetic rather than to please the convinced. 🟡
What the calculator actually shows: the payment is one-off¶
The architect's correction: "the payment is one-off, at the election." In the table above the assistant presented the rent monthly and thereby blurred two entirely different numbers. The difference is substantial and must be kept apart.
| Quantity | What it is | How much |
|---|---|---|
| Rent per person | how much flows past a citizen today; not what he receives | 76,177 ₽ a year (2024), about 380,000 ₽ over a five-year cycle |
| Dividend D = M × 1.5 × K | what a person receives — once, the day before the election (004.4.8) | at K = 1 %, about 11,000 ₽, roughly four working days of median pay (033c) |
| Efficiency bonus B | a share of budget savings, six months after the election and after audit (004.4.8) | unknown in advance by construction |
And here it becomes visible that the calculator computes something other than it seemed. The dividend is tied not to the rent but to median earnings: no parameter of the formula is controlled by the authorities holding the election (001b). The rent is the fund's source, the median is the meter. So "your share of the rent" and "your payment" are different numbers, and both must be shown without being mixed.
This also exposes a discrepancy the chapter already flags as a weak point (§8), which now becomes the calculator's principal subject. The architect's intuition is "a monthly salary or two pensions"; the formula at K = 1 % yields about a sixth of a monthly salary and roughly half an average pension. Both statements are true, because K is precisely what a referendum decides. For the payment to equal a month's earnings, K has to be of the order of six or seven per cent.
The arithmetic is open: at K = 1 % the payment is about 3 % of the rent accruing to a person over a cycle; at a month's-earnings payment, about 18 %. So the formula as it stands is deliberately conservative, and the rent sustains appreciably more. This is the argument for the changeable figures in §5: the K slider is not an interface convenience but the principal political question put on the screen. A person moves it and sees what the dispute is actually about.
Hence a correction to the architect's "a monthly salary or two pensions": those describe not what the formula yields today but what he wants K to be. Saying it that way is both more honest and stronger: not "you will be given a monthly salary" but "here is what flows past you, here is what is proposed to be handed over, and here is the slider that decides it."
Weak point. Median earnings were verified on 24 September 2026: by Rosstat the 2025 median is 65,307 ₽ a month (73,400 on the broader methodology), that is, about 784,000 ₽ a year and D ≈ 11,750 ₽ at K = 1 % — the repository's ~11,000 is right in order and slightly low [verified against publications]. The earlier caveat, for the record: the figure of about 11,000 ₽ is taken from 033c, and working backwards gives an annual median of about 733,000 ₽, which must be confirmed against Rosstat before any calculator is published. The size of bonus B is not estimated at all — it depends on budget savings that do not yet exist. 🟡
The third scale: what it costs the budget as a whole¶
The architect's correction: the comparison should be between the budget's total outlay on the payment and a budget line such as the State Duma. That is the right level, and it answers the objection that comes first: "where will you get that kind of money."
The architect's correction, and it halves the result. The first version computed the payment across all voters. That is wrong: not everyone takes the sum — those who choose exit do, and their share is the quantity being measured (004.4.3). The architect's estimate is 50–60 %. Elections also occur once in four or five years, so the annual figure comes from dividing.
The calculation. The electorate is taken as 110 million [an order-of-magnitude estimate, unverified 🟡], recipients as 50–60 % of it, i.e. 55–66 million. The cycle is five years. The arithmetic is shown so it can be checked.
| At K | Payment to one person | Total per cycle | Annualised |
|---|---|---|---|
| 1 % | 11,000 ₽ | 11,000 × (55–66 m) = 0.61–0.73 trn ₽ | 121–145 bn ₽ |
| ≈6–7 % (a month's earnings) | 70,000 ₽ | 70,000 × (55–66 m) = 3.9–4.6 trn ₽ | 770–924 bn ₽ |
The same against annual budget lines (at the midpoints, 133 and 847 bn ₽):
| Budget line | Per year | Dividend at K = 1 % | Dividend at "a month's earnings" |
|---|---|---|---|
| The State Duma | 14.84 bn ₽ | 9 times larger | 57 times larger |
| The entire Federal Assembly | 22.56 bn ₽ | 6 times larger | 38 times larger |
| Section "General state matters" | 2,251 bn ₽ | 5.9 % of it | 38 % of it |
| Oil and gas revenue, 2024 | 11,131 bn ₽ | 1.2 % | 7.6 % |
What this reads as, and it is the answer to "where from." The whole dividend under the present formula costs the budget about nine State Dumas a year — and a little over one per cent of a single revenue line. Even a payment the size of a month's earnings fits within eight per cent of oil and gas revenue and comes to a little over a third of the "General state matters" section.
And the correction exposes a property the analysis lacked entirely: the outlay is self-limiting, and limited in the right direction. The dividend's cost is proportional to the number who take it, and that number is the measure of dissatisfaction. A well-performing government pays less, because fewer choose exit; a badly performing one pays more. The budgetary load rises exactly when the government deserves it to and falls when it deserves the opposite.
This inverts the ruin argument: the budget is strained only by dissatisfaction, and the remedy for the strain is not to cancel the payment but to stop giving people cause to take it. No expenditure line today has that property.
The comparison with the Duma cuts both ways, and the second way must be stated honestly. It shows that the dividend is not small change: nine parliaments a year is real money, and the line "take it from the deputies and hand it out" does not add up arithmetically (see above). But it also shows that the money exists and sits in one line, rather than having to be scraped from the whole budget.
The fifth scale, stronger than the last: for companies the money exists¶
The architect's verdict on the previous section: "the comparison with luxury is not much good." Correct, and the reason is in its own weak point: every official car for a year is three per cent of the dividend. It is the very genre of outrage that divides into nothing (§3).
His replacement is more precise: find where the budget helps companies. "For them the money exists; for voters it does not." Here the sums are not three per cent but multiples of the dividend and, more importantly, it is the same kind of expenditure — money moved from the budget onto someone else's balance sheet with no goods received in return. Only the recipient differs.
For comparison, the dividend at K = 1 % per cycle: 55–66 m recipients × 11,000 ₽ = 605–726 bn ₽, midpoint ≈ 665 bn ₽.
| What the state pays companies | Sum | How many dividend cycles |
|---|---|---|
| Subsidising interest rates on preferential loans — compensating banks the rate differential on mortgages, SMEs, agriculture and industry, in the single year 2025 | 2,787 bn ₽ (8.8 % of all budget spending) | 4.2 cycles |
| — of which the overspend against plan, which passed with no public discussion | 1,000 bn ₽ (+57.7 % over plan) | 1.5 cycles |
| The damper paid to oil companies, 2024 | 1,815 bn ₽ | 2.7 cycles |
| The damper, 2022 | 2,171 bn ₽ | 3.3 cycles |
| The damper, 2025 | 881.8 bn ₽ | 1.3 cycles |
| Section "National economy," 2026, in one year | 4,770 bn ₽ (10.8 % of spending) | 7.2 cycles |
| State support for agriculture, 2025 actual | 665 bn ₽ | exactly one cycle |
| Share issue of the United Aircraft Corporation, approved January 2026 | 544.9 bn ₽ | 0.8 cycles |
| Cumulative state support for Rosnano, 2007–2018 | 404.6 bn ₽ | 0.6 cycles |
All figures [verified via publications citing Minfin, the Audit Chamber and Vedomosti; Minfin's primary files would not open from this environment 🟡].
The key figure here is the first, not the damper. Compensating banks for the rate differential in the single year 2025 was 2,787 bn ₽ — four times the entire dividend for the entire country over a five-year cycle. And the overspend deserves separating: the plan was exceeded by a trillion roubles, and it passed without public discussion. A trillion "turned up" within a year — a cycle and a half of dividends — and no referendum or explanation of its origin was required.
A coincidence worth remembering, because it explains itself. State support for agriculture in 2025 was 665 bn ₽. The dividend for the whole country over a five-year cycle under the present formula is also about 665 bn ₽. One industry for one year equals every citizen for five.
Why this beats both luxury and the deficit. Luxury is a moral argument — "they steal, and it is vile" — and it is small and does not divide. The deficit is the finance ministry's argument: an abstraction the person neither caused nor controls. A subsidy is a structural argument: nobody stole anything, the money was spent lawfully and by decision. So the dispute is not about honesty but about priority, and on that ground there is no answer: if a trillion over plan is found for banks within a year, "there is no money for payments to citizens" stops being an economic statement and becomes a statement about the order of the queue.
Hence the shortest formulation the whole section reduces to: the state already pays a dividend — its recipients are simply companies rather than voters. The protocol introduces no new kind of expenditure. It changes who stands first in the queue.
The honest half, without which the argument becomes demagogy. Subsidies have a stated purpose: cheaper mortgages, steadier fuel prices, a livelier industry. That is a choice rather than theft, and the argument must engage the choice rather than substitute an accusation for it. Three substantive caveats.
First, the damper is legally not an expenditure but a deduction against the reverse excise — it reduces oil and gas revenue rather than appearing among outlays. For the argument that is if anything a strengthening (the money fails to reach the treasury precisely because it goes to the companies), but "they paid it out of expenditure" must not be said — that is where one gets caught.
Second, part of the subsidy does reach people. A subsidised rate lowers the borrower's payment even though the bank is paid. How much of the benefit reaches the borrower and how much goes into the price of housing is contested, and need not be settled here; it is enough to note that the recipient of the payment is a bank, while the citizen benefits indirectly and did not choose it.
Third, the wealth fund's holdings in companies are not an annual payment but the value of holdings on a date, and cannot be added to expenditure. Russian Railways, with 722 bn ₽, is therefore left out of the table above, striking though the figure is.
Weak point. Minfin's primary documents (the list of tax expenditures, the damper table) would not open from this environment, so every row comes from publications citing them. Actual execution of the "National economy" section for 2025 was not found, so the 2026 plan is shown. The total of tax expenditures for 2024–2025 could not be confirmed; the only figure seen was 5,040 bn ₽ for 2023, and that is a forecast from a 2020 document rather than an outturn. 🟡
The fourth scale: not the deficit, but what people regard as luxury¶
(This section is kept as a unit of measurement — "how many dividends in one car" — but as an argument it is weaker than subsidies, see above.)
Two corrections from the architect, both methodological. First: comparing with the budget deficit is pointless — that is the finance ministry's argument, not a person's; the comparison should be with galling lines such as leasing cars for officials or maintaining the apparatus. Second: the budget is annual and the payment comes once per cycle, so the two must be brought to one basis.
Reduced to a five-year cycle: annual lines multiplied by five, the payment taken as it is.
| Line | Per year | Per cycle (×5) | Status |
|---|---|---|---|
| State procurement of passenger cars by all government bodies (4,400 contracts, 10,700 cars, average price 1.94 m ₽) | 20.8 bn ₽ | 104 bn ₽ | 2025, RTVI and Tenderscope's count from the contract register [verified; a journalistic aggregation, not a budget line] |
| Functioning of the President and his Administration | 32.9 bn ₽ | 165 bn ₽ | 2026, budget explanatory note [verified] |
| The Government apparatus | 13.0 bn ₽ | 65 bn ₽ | 2026 [verified] |
| Section "Mass media" | 146.3 bn ₽ | ≈544 bn ₽ | 2026; falls to ≈100 bn in 2027–2028, so the cycle figure sums the years rather than multiplying [verified for 2026–2028] |
| — of which RT ("TV-Novosti") | ≈28 bn ₽ | ≈140 bn ₽ | 2026 [verified] |
| — VGTRK | ≈25.6 bn ₽ | ≈128 bn ₽ | 2026 [verified] |
| Dividend at K = 1 % | — | 605–726 bn ₽ | 55–66 m recipients × 11,000 ₽ |
The first conclusion, and it answers "ruin" directly. Cars, the President with his Administration, and state media together come to 813 bn ₽ per cycle — more than the entire dividend at K = 1 % (605–726 bn). The cautious variant of the protocol is thus fully covered by three lines a person already regards as excess, with no hospital needing to be closed for it.
The honest half: at a month's-earnings payment (3.9–4.6 trn ₽ per cycle) those three lines do not cover a fifth of it, and reductions far larger would have to be named. Same conclusion as before: a low K is defended, a high one needs a separate conversation.
The exchange rate: how many dividends there are in one line of luxury¶
The comparison works harder in reverse — not "what the dividend costs" but how many people could be paid with the same money. At K = 1 % one dividend is 11,000 ₽.
| Expenditure | Sum | That many dividends |
|---|---|---|
| State media, per year | 146.3 bn ₽ | 13.3 m people |
| Every car bought by officials, per year | 20.8 bn ₽ | 1.9 m people |
| The President and Administration, per year | 32.9 bn ₽ | 3.0 m people |
| RT for a year | 28 bn ₽ | 2.5 m people |
| Chauffeured transport for one ministry, a single contract | 353 m ₽ | 32,000 people |
| Refurbishing one hall of Tatarstan's State Council — marble, Venetian plaster, premium jacquard seating | 288 m ₽ | 26,000 people |
| One VOYAH Dream minivan bought by the emergencies ministry | 9.0 m ₽ | 817 people |
This is the unit in which the conversation becomes intelligible without an economics degree. "One refurbished hall is the payment for an entire district town." "One nine-million car is eight hundred people." The same sums in pensions: the hall equals ten thousand average pensions, the car three hundred and thirty [average old-age insurance pension for 2026, 27,100 ₽, verified].
Why this beats comparing with the deficit. A deficit is an abstraction a person neither controls nor caused. A car, a hall and a television channel are things he sees and already has an opinion about. In this frame the protocol stops looking like a new expense and looks like what it is: a redirection of money already flowing.
Weak point, and a substantial one. Multiplying an annual line by five is crude: lines change, and media spending is planned to nearly halve after 2026, which is why its cycle figure sums the years. The car figure is a journalistic aggregation of the contract register rather than a budget line, and it is a lower bound: chauffeur hire and fleet maintenance are excluded. The budgets of the Presidential Property Directorate and the Federal Guard Service could not be verified at all — the first is not served on the open web, the second is classified. And above all, individual procurements are vivid but small: every official car for a year is 1.9 m dividends out of 55–66 m, about three per cent. A dividend cannot be built on luxury, and claiming otherwise is not allowed; luxury here is a unit of measurement, not a source of funding. 🟡
"This would ruin the budget" — the answer and its honest half¶
The architect names the main counterstroke: the authorities will press the line that this ruins the budget. The objection is predictable, and it has to be answered with figures rather than indignation.
What arithmetic answers. Under the present formula the dividend is 2.2 % of a single revenue line and 11 % of the section spent on general state matters. The state spends nine times the whole dividend on that section and does not call it ruin. So the dispute is not about the sum but about the direction: the same money going to procurement from friends is not considered ruin (§7). The question "does two per cent ruin a budget" is worth asking out loud — it refutes itself.
What the construction answers. The protocol does not demand a fixed sum regardless of circumstances. It demands that the deduction be named: "put defence spending into the algorithm — we will see what protection costs, and divide the remainder among us" (§7). Against an unspecified sum the ruin argument works; against a formula where every deduction is visible it turns into "here is what we deduct and why," which is precisely the conversation the protocol is after.
And the honest half, which we did not have until now — with the architect's caveat that the deficit is the finance ministry's argument rather than a person's, so in conversation with people it comes after the section above. The budget genuinely runs a deficit: 2025 revenue 37,284 bn ₽, expenditure 42,928 bn ₽, deficit 5,644 bn ₽ [Minfin, verified]. Against that:
| Variant | Annual cost | Addition to the deficit |
|---|---|---|
| K = 1 % | 121–145 bn ₽ | +2.1–2.6 % |
| A month's-earnings payment | 770–924 bn ₽ | +13.6–16.4 % |
From which follows a conclusion inconvenient for loud presentation but necessary: the arithmetic fully supports the cautious variant and supports "a monthly salary" only with a statement of what gets cut. Two and a half per cent on an existing deficit is noise the finance ministry would not notice; fifteen per cent is a visible fiscal decision requiring named reductions, and pretending otherwise is dishonest. Both figures are half those in the first version of this section, which computed across all voters rather than those taking the sum.
Hence a practical rule for any conversation about the protocol: start from a low K and show the slider, rather than promising a monthly salary. A low K is fully defended by arithmetic; a high one requires a separate conversation about what goes. This, incidentally, is the answer to why the repository's formula is deliberately conservative.
And finally. The claim that "paying citizens would ruin the budget" is itself a confession: it means the budget rests on citizens not being paid. Said out loud, it works against whoever says it — exactly as a ban does in §7.
Caveats without which the table misleads. The payment comes from a pre-formed fund (004.4.8) — it is not free, and it means that money is not spent on something else; the table shows scale, not the absence of a price. The efficiency bonus B is additional and not counted here at all. The number of recipients is doubly an estimate: both the electorate (110 million) and the take-up share (50–60 %) are estimates. On a four-year rather than five-year cycle the annual sums are a quarter higher. The take-up share is also not constant — it is the quantity being measured — so "annual cost" here means an average rather than a budget line. And the budget outlays are planned figures from the law, not actual execution. 🟡
A correction to the word "free"¶
The architect's phrase "two pensions free" is accurate as a sensation and inaccurate as a position, and the difference is expensive. "Free" reads as a handout — and triggers two ready-made objections at once: the paternalist "they will drink it away" (§7) and the envy of another's income examined in 056d.5b. The accurate formulation is "your share": not a gift but the distribution of what Article 9 of the Constitution already calls the foundation of the peoples' life. This is both truer and disposes of both objections before they arise: nobody envies his own share, and nobody asks whether its recipient will drink it away.
Weak point. The computation turns on what portion of the rent counts as distributable — a contested question where any value will be disputed, which is why changeable figures here are not a convenience but a condition of honesty. The app displays a number and does not prove that the number moves anyone; testing that is point 2 of §4. 🟡
6. Can a referendum application be filed in Russia: the procedure and where it breaks¶
The architect's question: can signatures be collected and an application for a referendum be filed. The check produced a more precise answer than expected, and refuted the premise §8 had been resting on.
The ban on budget questions was repealed¶
The assistant had assumed a question on distributing rent was prohibited by content: points 6 and 7 of part 5 of Article 6 of the constitutional law on referendums excluded questions on the budget, internal financial obligations and federal taxes. Those points were repealed — constitutional law of 24 April 2008 No. 1-FKZ; the current text reads "6) — 7) repealed" [verified against two independent publications of the text].
The prohibition did not disappear but moved into a more general point 10: questions "assigned by the Constitution and federal constitutional laws to the exclusive competence of federal bodies of state power." The wording is broader and therefore more flexible — but the law no longer contains a direct ban on a money question.
The Constitutional Court spoke in the same direction. Ruling of 21 March 2007 No. 3-P: the ban on questions concerning the performance and modification of internal financial obligations is constitutional only insofar as it concerns budgetary obligations as such, and does not imply a ban on referendum questions whose answers might change expenditure obligations beyond the term of the budget law [verified]. A distribution rule introduced for future periods does not fall under that description.
And there is a direct precedent. In August 2018 the Central Election Commission ruled by three decisions that questions on the retirement age complied with Article 6 (Nos. 174/1410-7 and 174/1411-7 of 8 August, No. 175/1416-7 of 10 August) [verified]. A question with colossal budgetary consequences passed the content filter.
A resource precedent does exist, and it is more useful than unwelcome¶
The first version of this section stated that no precedent on a resource question had been found. One was found, and it deserves honest examination.
CIK decision of 20 April 2005 No. 143/981-4 found fifteen of seventeen questions from a Communist Party initiative group non-compliant with Article 6. Among them question No. 9: whether "subsoil, forests, water and other natural resources, nuclear, thermal and hydroelectric stations, defence-industry enterprises, railways, high-voltage power lines and trunk pipelines must be exclusively in state ownership." The refusal was upheld by the Supreme Court (2 June 2005) and by the Constitutional Court (the same ruling No. 3-P) [verified against the full text of the court decision].
There were three grounds, and distinguishing them is essential:
| Ground | What the CIK said | Does it transfer to the protocol's question |
|---|---|---|
| Art. 6 part 7 — unambiguity | the question comprises several, no unambiguous answer is possible | no, if the question is single. The Communist formulation packed eight subjects into one sentence, from subsoil to pipelines |
| Art. 6 part 6 — no diminution of rights | the priority of state ownership diminishes the rights of other owners (Constitution arts. 8, 9, 35, 36) | no: the protocol does not touch ownership at all and diminishes nobody's rights — it distributes income the treasury already collects |
| Art. 6 part 5 point 6 — budgetary | "in substance this poses the question of nationalisation… permissible only on prior and equivalent compensation, and requiring expenditure from the federal budget" | lapsed: the point was repealed in 2008 |
What the precedent yields is not a prohibition but a specification. The only resource question the CIK has assessed was refused for being about nationalisation and for comprising eight questions at once. A question about distributing income is neither. Hence three requirements for the wording, which go straight into point 3 of the plan in §4:
- One subject. Not a list of industries but a single rule.
- Do not touch ownership. The question does not ask who owns the subsoil; it asks how the income the state already collects is distributed. This is exactly the distinction §7 of this chapter rests on: the protocol does not rewrite the Constitution, it puts it to work.
- Do not demand compensation for anyone's seized assets — that was the basis of the third ground, which is now repealed, but there is no reason to reproduce its logic.
The section's conclusion survives in refined form: the content filter kills not every resource question but a question about redistributing ownership. No precedent exists either for or against a question about distributing income, and it still cannot be asserted that such a question would pass. 🟡
For completeness: since 2008 the only federal question rejected by the CIK concerned accession to the WTO (decision of 13 April 2012 No. 118/917-6, upheld by the Supreme Court) [verified].
Where the procedure actually breaks¶
| Stage | Requirement | Deadline |
|---|---|---|
| Regional subgroups | at least 44 subjects (more than half of 89), each with at least 100 members, signatures in the registration protocol notarised | all subgroups within 2 months of the first one's registration |
| Notification and registration | notify the regional commission 5 days before the meeting; it registers within 15 days; the CIK checks the question against Article 6 within 10 | — |
| Registration of the initiative group | CIK, 10 days after the subgroups are assembled | — |
| Signatures | 2 million, no more than 50,000 per subject | 45 days |
| Verification | at least 40 % of signatures checked; 5 % defective means refusal | — |
| Appointment | the President, having first referred the papers to the Constitutional Court | — |
The minimum at the first stage is about 4,400 people across 44 regions, each before a notary, within two months — and that is before a single signature is collected.
But it breaks on one line rather than on the arithmetic. Part 18 of Article 15: once one subgroup is registered, the same commission may not register another with a formulation "the same in meaning or content" until the procedures conclude [verified].
Here the first version had the mechanism wrong, and the correction matters. The assistant wrote that in 2018 five competing groups blocked each other under part 18. Verification did not bear that out.
What actually happened [verified against the full texts of the decisions]: on 8 and 10 August 2018 the CIK, by five decisions (174/1410-7, 174/1411-7, 174/1412-7, 175/1415-7, 175/1416-7), found compliant with Article 6 the questions of regional subgroups — Moscow, Moscow oblast, Altai krai, Nizhny Novgorod and Vologda oblasts. Not initiative groups: a federal initiative group is registered only once 43 subgroups are assembled, and none has ever been registered under this law.
By late September there were about thirteen subgroups [second-hand 🟡]. On 17 October 2018 the CIK terminated the procedures on all five questions, and the formal ground was part 26 of Article 15: the two-month term had expired, counted from the registration of the first subgroup (Vologda oblast, 13 August).
So what kills it is not mutual blocking under part 18 but the clock. And it works more simply and more harshly: the two-month term starts with the first subgroup registered — anyone's. An early, uncoordinated or hostile registration burns the term for everyone working on the same question, and by the time a real campaign unfolds there is no time left.
Part 18 does exist and forbids registering, in the same subject, a subgroup whose formulation is "the same in meaning or content." But no source was found applying it as a ground for refusal, so mutual blocking cannot be presented as fact — it is an interpretation. 🟡
The federal verdict¶
In twenty-two years under the 2004 law not one citizen-initiated federal referendum has been held, no federal initiative group has ever been registered, and nobody has reached the signature stage. The CIK publishes no aggregate statistics; only this negative fact is firmly established. The known attempts are 2005 (the Communist Party), 2012 (WTO accession) and 2018 (retirement age).
This is the substance of §8, now with its precise cause: the problem is not that the question may not be asked but that the procedure is designed so that nobody reaches the question.
The local level: different, and with a working precedent¶
Regional and local referendums fall under federal law 67, and the thresholds are not comparable:
| Initiative group | Signatures | |
|---|---|---|
| Regional | 20 people | no more than 2 % |
| Local | 10 people | no more than 5 %, but at least 25 |
A content restriction exists there too — Article 12 part 8(g) bars questions on adopting or amending the budget and on financial obligations. But there is no ban on taxes, and part 9 states expressly: "the establishment of other restrictions is not permitted" [verified].
And here is the working precedent, worth more than everything else in this section. Since December 2003 4,844 local referendums have been held, almost all of them on self-taxation of citizens — one-off payments by residents into the local budget [second-hand 🟡]. Self-taxation may by law be decided only by referendum: a money question that the part 8(g) ban does not catch.
So a channel for a money question at local level exists and works in thousands of instances. Self-taxation is a referendum on how citizens pay into a budget; the protocol's question is its mirror — how a budget pays to citizens. Whether the mirror question survives the same filter nobody has tested, and it cannot be asserted. But it is the first procedure found in Russia that is not closed to a money question.
A timing caveat: federal law 131 is being replaced by the law of 20 March 2025 No. 33-FZ, in which the subject of a local referendum narrows from "questions of local significance" to "questions of directly providing for the population's life support" [verified; commencement dates and the transition period could not be confirmed 🟡]. The narrowing may close this channel too, and that must be checked before anyone relies on it.
And confirmation that procedures are written to fit the task¶
The 2020 vote on constitutional amendments was conducted not under the referendum law but under Article 2 of the constitutional amendment law of 14 March 2020 No. 1-FKZ — a self-contained procedure in twenty-six parts [verified]. The initiative lay with the President; there was no turnout threshold — amendments counted as approved on a majority of those taking part, whereas the referendum law requires turnout of more than half the registered electorate [both provisions verified].
So when the authorities want a vote, a new procedure is written — without a quorum and without the Article 6 filter. That is not an argument for the protocol but an exact description of what it deals with.
What follows for the plan in §4¶
Point 3 of the plan — a ready text of the question — becomes more concrete and cheaper than it looked. A point not previously there is added: test the local level. Self-taxation shows that money referendums are held in Russia by the thousand; the distribution question is its reverse in direction, and for that reason it is worth filing once in some municipality with rent of its own. The cost of the attempt is ten people and a lawyer; the result is informative either way — either the procedure is open, or a refusal produces a written ground that nobody has today.
Weak point. The CIK data on five competing groups in 2018 come from a secondary source and were not independently confirmed. No precedent on resource rent was found at either federal or local level, positive or negative. The figure of 4,844 local referendums is also secondary. And above all: all of this describes a procedure rather than a political reality — the fact that the content filter would pass the question says nothing about what happens to the people who file it. 🟡
7. Collect the signatures even if the deadline is burned: what is right here and where the danger lies¶
The architect's proposal: if the whole opposition unites, a non-political question will be signed by people within two months. And if the authorities file a fake application in advance to burn the clock, finish the collection anyway, doubling the number of signatures, and collect in the same token app. How would the authorities, looking at a doubled verified figure, say "you missed the deadline"?
Taken in parts: one is stronger than presented, one is right as presented, and one is more dangerous than it looks.
Right and underrated: a non-political question is signed differently¶
This is the core of the proposal, and the 2018 material supports it. Subgroups on the retirement age registered in Moscow, Moscow oblast, Altai krai, Nizhny Novgorod and Vologda oblasts (§6) — not a map of liberal capitals. A money question collects where a political one does not.
And non-political has a second, less obvious consequence: it makes unity technically possible. A joint campaign behind a common list of candidates requires everyone to accept somebody else's leaders; a campaign for a rule requires nobody to endorse anybody. A person who cannot stand the neighbouring column can still join a question about distributing income. That is precisely the property no item in the old repertoire had (§3), and possibly the only known way around what Gallyamov calls not creating losers (056d.5b).
Right, but for a different reason than stated: finish the collection — the product is not the authorities' embarrassment¶
The argument "how will they say you missed the deadline" counts on embarrassment, and they will not be embarrassed: terminating the procedures is one line in a decision, and in 2018 it was written without difficulty (part 26 of Article 15, §6).
But the collection should still be finished, and for a stronger reason: signatures here are a measuring instrument rather than a legal one. Legally they die with the deadline. Informationally they do not. Four million verifiable signatures tell each signatory that four million others signed, and that is exactly the common-knowledge mechanism 056d.7b rests on: Kuran and Chwe describe regimes falling not through persuading opponents but through opponents learning about each other.
And the comparison with 056d.7b favours signatures: a sign on a ballot is not counted, signatures are. This is the first instrument proposed anywhere in the repository that produces a number where there is nobody to produce one.
Hence the precise formulation: the collection is finished not to embarrass the authorities but to obtain a figure that can neither be drawn nor refuted. A refusal on deadline grounds does not destroy it — it publishes it.
Doubling the signatures does not work legally¶
One detail to dispose of at once. The law does not reward excess: at least forty per cent of signatures are checked, and five per cent defective means refusal regardless of the total (§6). Doubling also doubles defects in absolute terms. As a legal move doubling is pointless; as a measuring one it makes complete sense, and that is how it should be presented.
The architect's correction: there are no dissenters here, and that is a property of the design rather than rhetoric¶
The assistant wrote the preceding paragraph as though signing were self-evidently an act of dissent. The architect objected with a scene that settles it better than any argument:
"Where are the dissenters here? A fitter, Vasya, comes to the foreman for advice: I want to receive money at elections, he says, I don't go anyway. And how does the foreman, man to man, explain that going is an expression of disagreement with our successful government — in a country with the largest subsoil wealth per citizen and all that?"
The objection is right, and the assistant made an error of a particular kind: adopting somebody else's category. "Dissenter" is a classification the authorities assign; the content of the act does not match it. Vasya is not protesting, he is asking about money.
And the foreman scene is a working test rather than an illustration. The foreman is no random figure here: he is the transmission belt of corporate mobilisation, the very link that produced the main result in September 2026 (048i.5f, layer 2). To talk Vasya out of it he must say one of two things:
- "the state does not want to give you your share" — that is, repeat the protocol's argument in his own words;
- "asking for money is disloyal" — that is, admit aloud that loyalty consists in not asking.
Both sentences lose, and there is no third. This is a fork of the kind collected in 001c: a move exists, but every move worsens the position.
And the strongest part of the objection is its last words. "The country with the largest subsoil wealth per citizen" is not our thesis but the official one: the riches of the subsoil have been a matter of state pride for decades. So the question "where is my share" is asked from inside the official picture of the world rather than against it. Vasya is not subverting the propaganda; he is taking it literally. The boast cannot be withdrawn, and the question following from it has no answer.
Hence a rule extracted from this chapter into general form (023b.2) — the foreman test: a technique is sound if a local boss cannot explain to his own worker why not to do it. The ballot sign (056d.7b) passes for the same reason, being compatible with coercion. A signature for a money question passes it more cleanly still: the sign has to be explained, the signature does not.
But the risk remains, and it lies elsewhere¶
The correction removes the frame, not the danger. Three clarifications, and they matter more than the earlier wording.
First: prosecution follows the organisation, not the content. The foreign-agent and undesirable-organisation laws attach to whoever conducts rather than to what is asked. However innocuous the question, if it is collected by a labelled structure everyone who touched it is exposed. So the question's innocence protects Vasya and does not protect the campaign.
Second, following directly from the first: the defence here is the one named in §4 — having no owner. A collection without an owner has nobody to designate undesirable. This is another independent confirmation that the main line of defence is the structure rather than the wording.
Third: a list of millions of identified people is dangerous outside politics too. A signature sheet carries full name, date of birth, address, passport series and number. Such a database is valuable to anyone who obtains it — from fraudsters to anyone else — and that has nothing to do with the signature's political classification. The storage requirement stands in full: count anonymously, and either do not collect personal data at all or do not bring it together in one place (see above).
The architect's instinct to collect "on a blockchain token prototype" is right in direction and wrong in form: a public immutable ledger is the worst possible store for a list of dissenters. The opposite property is needed — proving the count without revealing identities.
And the repository has already solved that task: it is exactly how the protocol's open channel works (048i.2) — the token is anonymous, the sum of tokens is verifiable by anyone, and no identity is disclosed to anyone. So the instrument need not be invented; it needs to be taken from our own construction.
Hence the answer: two different instruments that must not be merged¶
| Legal collection | Anonymous count | |
|---|---|---|
| What is collected | signature sheets with passport data | anonymous tokens |
| What for | filing with the CIK | measurement and common knowledge |
| Volume | 2 m by law | whatever is collected |
| Risk to a participant | high and personal | near zero |
| What a refusal yields | the end of the procedure | publication of the figure |
They must not be merged, and here is why: the moment an anonymous count is turned into legal signatures it stops being anonymous. An anonymous figure cannot be filed with the CIK — and need not be: the legal route is closed by the deadline anyway, and the measuring route does not suffer for it.
The practical consequence for the §4 plan: the anonymous count is something that can be built today, without initiative groups, notaries or deadlines, using the same app that computes the share (§5). The legal collection is something to undertake only when ready to answer for signatories' safety, and that question is settled before the first signature rather than after.
Weak point. "If the whole opposition unites" is an assumption rather than a plan; §3 shows it did not happen once in a full cycle, and the question's non-political character makes unity possible without securing it. An anonymous count has no legal force at all and is easily dismissed as inflated — its credibility rests solely on the openness of the counting scheme, which requires publishing all tokens as in 048i.2. And neither instrument answers what happens to the organisers: measurement protects participants, not those conducting it. 🟡
8. "The thirteenth salary" and the word "freebie": the best unit and a dangerous word¶
The architect's remark completing the previous scene: the poor work at a factory for forty thousand a month, and here they are offered a thirteenth salary, lawfully, from the budget. "For Vasya, and even for the foreman, and for everyone else, it is a freebie. Russians and freebies…"
Two different things here, and separating them is obligatory: one is the best unit of measurement found anywhere in this chapter, the other a word that must not be spoken, though it describes reality correctly.
"A thirteenth salary" beats every earlier formulation¶
The chapter's search for units ran thus: roubles a year (mean nothing), "a monthly salary" (means something but sounds like a promise), "two pensions" (the same), "your share" (true but abstract). The thirteenth salary beats each, for four reasons at once.
It is familiar as an institution rather than as news. An annual bonus of roughly a month's pay is a concept known to anyone who has worked in a factory; nobody needs it explained.
It is a settlement rather than a handout. A thirteenth salary is not begged for — it is received at the end of a period. That is exactly the frame §5 demanded in place of the word "free": not a gift but something due. And unlike "your share" it does not require first agreeing with an argument about the subsoil — it is intelligible before any argument.
It is periodic and tied to the end of a cycle — as is a payment arriving the day before an election (004.4.8).
And it is wholly non-political, so it passes the foreman test (§7) without a word of explanation.
But the arithmetic must be stated honestly¶
A thirteenth salary for a worker on forty thousand is forty thousand. The formula at K = 1 % yields about eleven thousand — roughly a quarter of a thirteenth salary. A full one requires K of the order of six or seven per cent (§5).
So "a thirteenth salary" may be said only together with the slider: under the present cautious formula it is a quarter, and what it becomes is decided by a referendum. Promising a full one without naming K is precisely what §5 warns against.
A second caveat, also to be said aloud: the payment comes once per cycle, not once a year. Forty thousand once in five years is about seven hundred roubles a month if spread out. But spreading it out is exactly what must not be done, and therein lies the point.
Why a lump sum is the mechanism rather than a shortcoming¶
For a person on forty thousand a month, eleven thousand in one sum and a hundred and eighty roubles a month sixty times over are entirely different events, though the money is the same. The first is something one can act on; the second is an imperceptible increment that dissolves into tariffs.
This is the financial judo of §5: the sum works not because it is large but because it arrives whole and on a particular day. Spread out, it would stop working altogether — which is why the lump must be defended as a design feature rather than apologised for.
"Freebie": an accurate forecast of reception and an unusable formulation¶
The architect is right in the observation: it will be received exactly so, and that applies to the foreman too — he is a recipient as well. Which incidentally closes the foreman test (§7) harder still: the foreman not only cannot explain why not, he is owed the same money, so he has no personal interest in dissuading anyone.
But as a formulation the word is unusable, and this is not a matter of taste.
"Freebie" means unearned. Saying it hands an opponent all three ready-made objections listed in §7: the paternalist "they will drink it away," the charge of bribing voters, and the dismissal as populism. Worse, it directly contradicts the frame "your share": one cannot simultaneously say "this is yours by right" and "this is a freebie" — the second sentence cancels the first.
And there is a third argument, from 056d.5b: in the limited-good worldview another's unearned gain provokes envy, while one's own does not. Calling the share a freebie moves it from the second category into the first and turns against us the very mechanism 056d.5b shows working in our favour.
Hence a simple, hard rule: "freebie" is how it will be understood and never how we speak. We say "a thirteenth salary" and "your share."
And the architect's correction, which removes half the worry: "Vasya will say that to the foreman without us." Right, and confirmed by the same 056d.5b. A freebie is what somebody else receives; one's own is what is due. A person to whom money is owed produces the legitimation himself, unprompted — he needs none of our vocabulary to tell the foreman "this is mine."
From which follows an unexpected argument for universality (004.4.3c): while everyone receives, the word "freebie" has nobody to attach to. It requires someone who received undeservedly, that is, an exception to the general rule. A universal share creates no such addressee; a payment to a selected group creates one at once, and 056d.5b shows on the war participants' example how quickly.
Weak point. The thirteenth salary is a custom rather than a legal right, and in today's Russia it is far from universal; younger workers may have no such experience at all, in which case the unit fails. Forty thousand is the architect's example rather than a measured median — the Russian median wage used in this chapter's arithmetic is still unverified (§5). The difference between a lump sum and a spread payment rests on a laboratory-measured dependence on the reference point, never replicated for Russia. 🟡
9. What happens to pro-government bloggers: their position is worse than the foreman's¶
The architect's question: what effect will there be on Z-bloggers and pro-government opinion leaders generally. Examined by the same method as the foreman scene (§7): not "what will they think" but which moves are available to them.
Two moves they do not have¶
They cannot say "the wealth is not yours." The thesis that Russia is the world's richest country in subsoil is theirs rather than ours: it has been a matter of state pride for decades and an anchor of the sovereignty argument (§7). Abandoning it for the sake of one dispute is impossible — too much else rests on it.
They cannot say "let the state keep it." That sentence requires conceding that the state and the people are different recipients — saying aloud the very distinction pro-government commentary exists to refute. Uttering it, a person formulates the liberal thesis of appropriation with his own hands.
It is the foreman's fork, only heavier: the foreman has no audience, a blogger does, and the audience remembers.
The move they do have: postpone¶
One option remains — "the idea is right, but not now": a war, a besieged fortress, no time for dividing. This is the third of the arguments listed in §7, and the only one open to them.
And that is better for us than rejection. Whoever postpones has already conceded the substance: he has confirmed that the share is due and disputes only the timing. The thesis becomes legitimate inside a loyal audience, and by a person it trusts. Timing arguments have the property of expiring: "not now" sounds worse in two years than today, and whoever said it will owe an explanation of why it is still not now.
Appropriation is a trap rather than a threat, and the repository already says so¶
The assistant called appropriation the principal danger. That is an error, made worse by contradicting his own repository — the architect said so directly: "re-read the repo, there is a trap there."
In the fork catalogue of 001c the two lines stand side by side:
| Player | The fork | Where |
|---|---|---|
| A heavyweight interceptor | seize the theme and work on the first mover's monument; leave it alone and the underdog keeps it | 047.3 |
| The incumbent | copy the idea and launch the audit that destroys him; do not copy and leave the underdog something that cannot be seized | 042.74.3, 030.14 |
The mechanism of the first line is stated in 047.3: authorship of the first move is inalienable — Brexit is stuck to Farage, not to Cameron who called the referendum. Everyone who subsequently names a sum works on the first mover's monument.
The second line is the one the assistant missed entirely, and it is decisive. To pay a share one must state how much rent there is, what is deducted, and among how many it is divided. That is an audit of the rent, performed by the state upon itself. Afterwards the gap between "how much there is" and "how much was given" becomes a public number, and the size chosen has to be defended aloud. Appropriation does not silence the question — it poses it in countable form for the first time.
Where the trap springs and where it does not¶
The distinction is single and precise: the trap springs on the wording, not on the payment.
A one-off payment before an election with no framing at all is not appropriation of the protocol but what already happens: money is handed out, no arithmetic is named, no question arises. Nothing new.
But the moment "your share of the subsoil" is uttered, the counting starts: a share of what, divided among whom, over what period. Those words cannot be half-taken — they drag a number behind them. This is precisely why §8 requires saying "your share" and "a thirteenth salary" rather than "a payment": the wording is the thing that cannot be appropriated in part.
Who pulls the trigger: there are many of them and they need an audience¶
The architect names the mechanism: "the dumbest bloggers will go there for the hype, because there are many of them and they need an audience — they will go and check."
That is an exact description of how an idea spreads without a champion. A major pro-government commentator weighs consequences and will probably choose "not now" (above). But the attention market is not made of major figures: hundreds of small channels compete for the same audience, an anti-elite money topic pays them immediately, and strategic consequences do not concern them — individually they have none.
And "they will go and check" is the key to the whole section: they check with the calculator (§5). Whoever came for the hype opens the count, gets a number about himself, and can no longer not know it. The calculator here is not agitation but the trap's trigger: it turns a borrowed slogan into personal arithmetic, and personal arithmetic does not fold back into a slogan.
What remains a genuine risk¶
Not appropriation but its degenerate case: a payment with no talk of a share — money before an election, not a word about rent. It springs no trap and takes nothing from the protocol.
The practical requirement therefore stands, on a different ground: the formula and the counter are inseparable from the payment not because the idea would otherwise be stolen but because a payment without them is not a weak version of the protocol but simply not it. The test is the same: who sets the size?
A handout and a share are different acts, not different words¶
The assistant wrote above that a payment without framing is "exactly what happens now," thereby reducing the difference to wording. The architect objected, and the objection goes deeper than anything preceding it:
"That is a handout, and it is humiliating. People want the bosses to be obliged. They want to make the bosses feel equal — that he is the authority here. And here there is a pretext to take back what is his, lawfully. These are not the same thing at all."
Right, and the error was not in a word but in the analysis: the same money transferred two ways creates opposite relations.
| A handout | A share | |
|---|---|---|
| Who acts | the boss gives | the boss is obliged |
| The recipient's position | supplicant | creditor |
| What is owed afterwards | gratitude | nothing |
| Can it be withheld | yes, that is the giver's right | no, that is a breach |
| What the act confirms | the hierarchy, and confirms it anew every time | an obligation |
The repository has named this position already — §7 describes the present arrangement as handing out "crumbs from the table" on a "say thank you" basis. But until now it was presented as rhetoric. In fact it is a mechanism, with three properties money as such does not have.
First: the addressee of the feeling is the boss, not the citizen. The architect puts it exactly: "make the bosses feel equal." This is not about the recipient's self-perception but about placing the other party in the position of a debtor. That is a motive of a different kind, and considerably stronger than self-interest.
Second, and practically important: this motive does not diminish with income. §5 notes that eleven thousand makes no difference to a Moscow professional, who would decline the sum. But "the bosses are obliged" is not about the sum. It works on anyone who has ever been made to say thank you, and therefore reaches where money does not. This is a second payload of the dividend the repository has underrated.
Third: a lawful pretext is a third option where there are usually two. Asking is humiliating and rebelling is dangerous; "taking back what is yours, lawfully" is neither. The repository already put this in §7: "you wrote yourselves that we are the beneficiaries — here is the till, ring it up." It is now clear that this is not a quip but a description of the only available position in which a person is neither supplicant nor rebel.
Indirect confirmation is self-taxation (§6): thousands of local referendums in which people vote to pay themselves. The money flows the other way and it still happens — because they decide. What works is not the direction of payment but whose decision it is.
And hence the basis for the rules about words¶
The bans in §5 on "free" and §8 on "freebie" were derived tactically: those words invite ready-made objections. They now have a deeper basis: both words return the recipient to the supplicant's position, abolishing precisely what the construction exists for. A handout called a share remains a handout; a share called a freebie becomes one.
And the test "who sets the size" acquires a second function. It separates the protocol from a forgery — already recorded. But it also separates a right from a favour: a payment whose size is set at discretion is humiliating regardless of its name and magnitude. Worse, it is worse than a plain handout, because it adds disappointed expectation to dependence.
So the formula's independence (001b) is not only protection against manipulation. It is what makes the payment a right rather than a favour. One element of the construction carries both functions, which is worth remembering in any conversation about "simplifying" the formula.
Weak point. Self-taxation is offered as an analogy and a weak one: there people pay themselves and the sums are small. How far "the bosses are obliged" outweighs financial interest has not been measured for Russia or, so far as the assistant knows, anywhere. And the argument has a reverse side worth holding in mind: the creditor's position is pleasant only while the debt is paid; non-payment of what is owed by right is felt more keenly than the absence of a handout. 🟡
The uncomfortable part: their audience fits better than the liberal one¶
This must be said plainly, because it is what the analysis yields.
The rent question is built as anti-elite, pro-people and sovereign — and requires no liberal premise whatever. It is not about free speech, not about the war, not about the West. The material in 056d.5b shows that the energy of "they are feasting in the rear" is live in that milieu and currently aimed at the war's participants; the question of a share lands on it without an adapter.
And the scene in §7 says the same from the other side: Vasya is not a liberal. He is a worker on forty thousand who wants what is due to him. He needs neither the "party of the big cities" (§3) nor its language for that.
From which follows a conclusion uncomfortable for both sides: the protocol may sit closer to the audience pro-government bloggers speak for today than to the audience currently discussing it.
And the risk that comes with that audience¶
It is named in 004.4.3c and acquires a concrete carrier here. A milieu for which "take it from the foreign agents" and "those who left do not deserve it" are natural formulations will demand exclusions first of all. And universality is load-bearing: a share made conditional becomes a benefit that somebody awards.
The argument against exclusions stays mechanical rather than moral: any exclusion creates the office that decides whom to exclude — precisely the office the protocol abolishes. That argument works identically in any audience, which is its chief merit.
Weak point. No pro-government opinion leader has publicly addressed the rent question, so this section enumerates available moves rather than describing observed behaviour. The fork "they cannot abandon the wealth thesis" assumes a commentator is bound by his earlier words — a weak assumption, since he is bound only as far as the audience remembers. And judging appropriation the likeliest outcome is the assistant's estimate with no precedent behind it. 🟡
10. Three different platforms, not one¶
The architect's correction to §9: "you are confusing them — what the authorities say, what the opposition on YouTube says, and what a permitted opinion leader says in Telegram channels."
The assistant did indeed collapse them into a single figure, the "pro-government blogger," and thereby spoiled the analysis: these three positions have different constraints, different audiences and different available moves. They must be taken separately.
| Official authorities | Opposition on YouTube | A permitted opinion leader on Telegram | |
|---|---|---|---|
| Where | television, statements, state media | blocked platforms, a foreign-agent label | inside the country, no VPN, no label |
| What can be said | almost nothing: any word about a share legitimises the question | anything | almost anything, as long as it is not directly political |
| Whom it reaches | everyone, with zero credibility on this topic | the already convinced (§3) | the mass audience |
| Best move | silence | say it | pick it up if the topic gets views |
And a fourth row the first version lacked — private channels and reposts; it appears below, because that is a transmission medium rather than a platform.
First correction to §9: the authorities' best move is silence, not "not now." The assistant attributed the besieged-fortress argument to them, but uttering it concedes that there is something to divide. The official response to a successful money question is not objection but the absence of a response; objection begins when silence is no longer possible.
Second: opposition YouTube can say it but cannot spread it. And not only because of blocking. A labelled source's touch makes a topic toxic for everyone else — exactly the mechanism that killed Smart Voting (§4): the app was not what was blocked, the label attached to the organisation and through it to everything connected. So uttering the idea in that channel is not free: it may close its entry into the others.
Third, and this is the main omission: permitted Telegram channels are the only bridge. They are inside the country, unlabelled, reachable without a VPN, and they compete with each other for the same audience. An anti-elite money topic pays them immediately, and individually they have no strategic consequences to weigh — the very mechanism the architect described in §9 ("there are many of them and they need an audience"). The idea's route to a mass reader runs through them rather than through opposition platforms.
Correction of 01.10.2026 to the second and third (§13l). Both were written on the model of Smart Voting, and the model does not fit: there the label hit the content — the instrument was for the opposition's candidates. The content of the money question is the voter's own money, and a label does not stick to it; the Bonya case showed that an attack on the person does not refute the sum and produces a second wave (§13g). The architect: "the protocol does not care who raises it". So "not free" and "the only bridge" are to be read as a judgement about speed, not about admission: a labelled source may hold back the pick-up, it cannot close the entry, and any channel may raise the question.
A fourth platform, larger than the previous three¶
The architect adds what the table lacked entirely: "and there are many ordinary private channels, and sometimes Russians post, not knowing better, what they take to be permitted by the authorities."
This is not a fourth producer but a transmission medium — and it is where everything actually happens. The three positions above produce something; private channels, chats and reposts carry it, and in volume they exceed all three together.
This medium has a property none of the three has: it is itself the filter. An ordinary person forwarding a post unconsciously checks it against "is this allowed?" — and that is the most widely applied censorship in existence, because everyone applies it constantly. If content has passed through it, it has passed the main barrier.
And here the question of a share passes easily: it reads not as politics but as talk about money — the most permitted genre in Russian-language private channels.
But "not knowing better" must be examined honestly, and here the assistant has an objection. The phrase allows two senses: either the person is mistaken and at risk, or the content genuinely is not prohibited and his judgement is correct. The difference is fundamental, and whether this channel can be relied on depends on it.
"I thought it was allowed" is not a defence, and people in Russia are prosecuted for reposts. So the rule must be hard: a technique is sound if the forwarder's judgement is correct, not if it is mistaken. An idea that spreads because people misjudge the risk uses them as cover; an idea that spreads because it genuinely is not prohibited takes nothing from them.
For the question of a share the second appears to hold — but verifying that is not the forwarder's job. Hence a requirement on form: the material must contain nothing that makes forwarding risky — no links to labelled sources, no mentions of organisations, no assessments of the authorities. Only the rule, the number and the calculator (§5). Everything else adds risk to someone who chose nothing.
The architect's reframing: this is not opposition, this is for the people¶
"A referendum to hand out money to the people and to yourself is not opposition at all, it is for the people."
This is the most precise statement of positioning anywhere in the chapter, and it explains why the preceding filter lets it through.
Opposition content invites you to be against something — and requires courage. Content that is "for the people" invites you to be for yourself — and requires nothing.
And crucially: "for the people" is the authorities' own register. State rhetoric is built on it; a phrase uttered in that register is indistinguishable by ear from television. Which is why it passes self-censorship: a person forwards not a protest but something that sounds like the state's position with a question attached to the end.
This continues §7 from the other side. There it was shown that the question is asked from inside the official picture of the world (the subsoil's riches being an official thesis). Here it becomes clear that it spreads for the same reason: the forwarder does not feel he is doing something forbidden, because he is not.
And the architect's observation that changes the distribution arithmetic¶
"People watch quietly and know everything — even what the authorities have de facto banned."
If so, the blockade does not prevent knowing — it prevents citing. A person may be well informed and still unable to name the source: citing a labelled one draws the label onto himself.
And hence something unexpected: the blockade itself produces the condition §4 calls the main defence. An idea everyone knows and nobody attributes has no owner — there is nobody to designate undesirable because there is nobody to point at. What is intended as isolating the source works as detaching the idea from the source.
The practical consequence for the §4 plan changes point 5. It previously read "spread through those who cannot be struck off." More precisely now: the idea must be built so that it can be retold without a citation. That is a requirement on form — a short rule, a countable number, a calculator that works by itself (§5) — rather than on a list of allies. What gets retold is what needs no authority in order to be checked.
Checked against figures: half confirmed, half broken¶
The assistant requested data on three of this section's assumptions. The result is mixed, and the worse part must come first.
Broken: the bridge called "the only one" no longer exists. Telegram is being blocked in Russia — throttling from 10 February 2026, and by 10 April blocking reached 95 % on an independent project's measurements [second-hand 🟡]. Telegram's average daily audience in August 2026 was 33.8 m, down a third from April; WhatsApp 22.6 m, down a quarter. And Max — 69.9 m [verified against a publication of Mediascope data]. By monthly audience Max overtook Telegram back in June 2026: 86.3 m against 75.7 m [verified].
So the mass medium §10 called the bridge is being dismantled as we write, and the audience is moving into a messenger controlled by the state. The claim that permitted Telegram channels are the only bridge was true when written and is already false.
The architect's methodological correction, and it strikes at the foundation¶
The assistant cited two survey figures: 40 % VPN users and the absence of opposition bloggers from a most-popular list. The architect objected:
"Why are you giving me statistics from Russia, where surveys are answered under fear? They are irrelevant. Where do millions of YouTube views come from if people are not circumventing? Where do the Z-channels' views on Telegram come from? VPN reach has been 80–90 % for a long time."
The objection is methodologically right, and the assistant fell into an internal contradiction. The whole chapter rests on Kuran: in regimes of this type people systematically conceal preferences, which is exactly why 056d.7b is built around giving them a way to report the truth anonymously. Leaning after that on a survey of those same people is applying an instrument whose uselessness one has just demonstrated.
The refinement needed here: the distortion is not uniform. "Do you use a VPN" is less dangerous than "do you watch so-and-so," because using a VPN is not prohibited to a citizen. But the direction of distortion is the same in both cases — downward. So 40 % is a lower bound rather than an estimate, and the absence of a named foreign agent from what people tell a stranger on the telephone says almost nothing about his audience.
Hence a rule for the whole chapter, to be held firmly: survey data from Russia do not belong here except marked as a lower bound. Its general form is recorded in 023b.3 — it is about fear, not about Russia. What serves is behavioural data — traffic measurements, metered panels, view counts, blocking. They do not ask a person, they observe him.
What survives the clean-up¶
Separating the sources shows that the section's load-bearing part was behavioural and stands.
| What | Type of data | Does it stand |
|---|---|---|
| Telegram blocking at 95 % on measurement | network measurement | yes |
| Audiences of Telegram, WhatsApp, Max (33.8 / 22.6 / 69.9 m) | metered panel | yes |
| YouTube daily reach under throttling | panel | yes |
| 40 % use a VPN | survey | only as a lower bound |
| Katz absent from the popularity list | survey | withdrawn as evidence of reach |
| 14 % for opposition channels against 44 % pro-government | survey | withdrawn |
And the architect's behavioural argument is stronger than the withdrawn figures. YouTube has been throttled since August 2024, and its daily reach in the first half of 2026 grew by 4 % [panel measurement, verified]. A platform being strangled for two years is growing — that is circumvention observed rather than declared. View counts on channels living on blocked platforms read the same way: a meter counts them, not a respondent.
But the assistant cannot confirm the 80–90 % figure, and taking it on trust would be the same error as taking 40 %. The honest formulation: the share using circumvention is above forty per cent, the true value is unknown, and behavioural data indicate mass use without yielding a percentage. What would settle it: VPN app download statistics, carrier data on the share of traffic over known circumvention protocols, or a platform's reach before and after blocking on a single panel. None of these was found in open access.
And the main consequence, which is about the protocol rather than the statistics¶
The architect's objection is not a remark about source quality but an argument for the construction itself, and it should be recorded as such.
If surveys do not work in such a country, then neither do ratings, nor "support for the course," and no sociology will say how many people are against. The protocol answers exactly that: it replaces asking with observing an action, and makes the action monetary, because money is collected even by those who would not answer a question. The architect put it earlier and more sharply: "surveys are not comparable to money in accuracy at all" (004.4.3b).
By leaning on a survey the assistant applied here precisely the instrument the protocol is built to replace.
What follows, and the conclusion is stronger than before¶
The picture has worsened, but worsened selectively — and it has eliminated every option but one.
If the mass medium is moving into a messenger owned by the state, then content that lives only in opposition channels does not reach a mass audience at all — neither by the bridge nor around it: circumvention gives access, not attention.
One type of content remains capable of existing inside a state platform: that which does not read as oppositional. Which is exactly what this chapter arrived at independently — "not opposition, for the people" (above), a question asked from inside the official picture of the world (§7), material forwarded without any sense of transgression.
From a tactical advantage this becomes a condition of existence. The argument used to run: being non-political helps distribution. Now: being non-political is the only condition under which distribution is possible at all, because every other channel is either closing or unread.
And hence a practical consequence for §5 worth recording outright: the calculator must be built to work and spread inside the state messenger — as a link, an image and a number forwarded without explanation and without a source, rather than as an app from a store or a clip on a blocked platform.
Weak point. The observation that "people watch and know everything" is the architect's testimony rather than a measurement; the share of Russians consuming blocked content has not been verified by the assistant, and the whole calculation depends on it. The table of three platforms describes constraints rather than observed behaviour: none of the three positions has publicly reacted to a rent question. And the conclusion that the blockade produces ownerlessness holds only where circumvention is widespread; where it is not, the blockade works as intended and produces nothing useful. 🟡
11. Does the opposition in exile have better ideas: an honest survey¶
The architect's question is direct: judging by how the opposition prepared for and ran these elections, are there better and more realisable ideas than the protocol right now?
Answering requires a survey rather than a confirmation. Below is what is actually on offer, assessed by two separate measures, because they must not be mixed: does it give an apathetic Russian a personal reason, and is it realisable today.
| What is offered | A personal reason for Vasya | Realisable today | What the cycle showed |
|---|---|---|---|
| Corruption investigations | no: outrage does not divide per person — the palace is 684 ₽ once (§5) | high, it is being done | a full cycle with millions of viewers produced zero districts (§3) |
| Smart Voting and booth tactics | no | nil: the class of techniques ends at a number the commission writes, and in September 2026 it stopped being embarrassed | went to zero at once |
| Sanctions and external pressure | no: the addressee is Western governments, not a voter | high, it is being done | no effect on the internal picture; and its touch makes a topic toxic (§10) |
| Work with emigrants | yes, but for those who already left | high | foreign precincts routed the party of power (§1) and nothing changed inside |
| The anti-war agenda | no: it requires courage and there is no valve (§2) | high | no protest voting occurred |
| Wait for a transition, prepare cadres and programmes | no | not a mechanism but a timeframe | — |
| The protocol | yes, and it is the only case in the table | the end point is closed, the intermediate ones open (see below) | never tried |
First conclusion: the slot really is empty, and that is not a compliment¶
By the first measure the answer is unambiguous: not one of the offerings even attempts to give an apathetic person a personal reason to turn up. Everything listed is procedural ("how to vote"), moral ("against the war," "they steal") or external (sanctions). Nobody asks "why bother at all," and this was established by survey rather than by argument (§3).
So to the architect's question in its literal form the answer is no, nothing occupying that slot is currently on offer. But it does not follow that the protocol is better than everything: it follows that it is alone in its column, and the column may be the wrong one.
Second conclusion: there is a serious rival, and it is a different theory rather than a different tactic¶
It must be named, or the survey is dishonest. Regimes of this type fall more often from elite splits than from pressure below — that is the main line of comparative transitology, and it is held by Shulman, on whom the repository relies elsewhere.
On that theory the right programme is not talking to Vasya at all but working on the apparatus: guarantees for non-security officials, personal forks for those with something to lose, preparing conditions in which leaving the system stops being suicide. Under that theory the protocol is not the best idea but a secondary one, addressed to people whose behaviour, on that theory, decides little.
That assessment is withdrawn at the architect's instance, and rightly — see §12. In brief: an elite split fails not on dissatisfaction but on coordination, and coordination requires a public pretext. Mass support is not an alternative to the split but its input.
One partial qualification in the protocol's favour: it worsens the apparatus's position regardless (048i.5g — turnout stops being a commodity and a local boss has nothing to sell upward), so it works along the elite line too. But it works there more weakly than targeted work with elites, and there is no reason to pretend otherwise.
Third conclusion, the least welcome: on realisability the protocol loses to almost everything¶
Investigations, sanctions and work with emigrants are being done today. The protocol needs adoption, adoption needs a referendum, and the procedure is closed by clocks and subgroups (§6). By this measure it is among the worst in the table.
What is realisable within it is narrow: the calculator (§5), a trial local referendum (§6), a ready text of the question, an anonymous count (§7). That is not a programme but four cheap actions.
Two corrections from the architect to the table¶
First: "low is low, not nil; and the discussion itself is already a phase transition."
The objection is right, and the assistant substituted one measure for another. Realisability was measured on a single criterion — will a referendum happen — that is, as yes or no. But with the protocol the intermediate states are results in themselves, not failed attempts at the final one:
| State | What it produces | Permission required |
|---|---|---|
| The topic is discussable in money terms | common knowledge: each learns what others think (056d.7b) | no |
| The calculator works | a personal number the person computed himself | no |
| An anonymous count is gathered | a figure that can be neither drawn nor refuted (§7) | no |
| A local referendum is filed | either a written ground of refusal or an open procedure (§6) | ten people's application |
| A federal referendum | adoption of the protocol | closed (§6) |
Only the last line requires what is unavailable. The four preceding are states in which the country is no longer what it was, and none is cancelled by refusal at the last.
"Phase transition" is the right term, and the chapter already has the precedent: the independent count in East Germany in May 1989 changed not a single result and changed what people knew about one another (056d.7b). Measured by "did the vote happen" it failed; measured by consequences it did not.
So the table's row should read differently: not "low realisability" but "the end point is closed, the intermediate ones are open and valuable in themselves."
Second, and it inverts the frame: "Vasya needs this more than the opposition does, he is poorer."
Here the assistant was wrong systematically, across the section rather than in one line. The analysis proceeded as though the protocol were the opposition's programme: what it might adopt, where to spend resources, how this compares with its other ideas. But the dividend stands in different relations to the two:
- for the opposition a share is an instrument: a way to bring a person to the polls;
- for Vasya a share is the object: money he needs.
And §5 already says this arithmetically: the sum changes everything for a nurse and nothing for a Moscow professional. So the protocol's natural carrier is not whoever would use it as a means but whoever it is the point for.
Hence a consequence the chapter lacked: the protocol need not run through the opposition¶
If it is needed more by the poorest, the opposition's agreement is a convenience of distribution rather than a condition of existence. And §10 showed that a labelled source's touch additionally costs: it makes the topic toxic for every other channel.
Four facts established in this chapter add up to one:
- §5 — the sum means more the poorer the recipient;
- §7 — the technique passes the foreman test, requiring no political position;
- §10 — an opposition channel does not distribute but infects, and the mass medium is private channels;
- §9 — the idea spreads through small actors who need audience, not through a champion.
All four say the route bypasses the opposition.
One thing it is nonetheless needed for, and this must be stated precisely. Somebody has to build the calculator, draft the question, launch the count. That takes resources and organisation, which Vasya lacks and the opposition has. So its role is builder rather than carrier: make it and let it go, do not brand it and do not campaign on it.
That is exactly the ownerlessness requirement of §4 and §10, now arrived at from a third direction.
Hence the precise answer¶
"The protocol versus the alternatives" is doubly the wrong framing. First, the protocol does not replace the repertoire, it occupies its single empty place. Second, and more importantly, it is not the opposition's programme at all: it is addressed to whoever needs it more, and the opposition is assigned a narrow role in it — to build and not to appropriate.
And the argument for it is not superiority but expected-value arithmetic: the expensive things were tried for a cycle and yielded zero; the cheap and unoccupied one has never been tried once. The four actions of §4 together cost less than one investigation, and any of them yields knowledge nobody has today — including negative knowledge.
Weak point, applying to the whole section. The table was assembled by the assistant from open material and the three September 2026 analyses; none of those listed has described their strategy in these terms, and imputing aims from outcomes is unreliable. The "personal reason" assessments were measured for no item in the table, the protocol included. And most importantly: the conclusion that the slot is empty comes from surveying what the assistant knows rather than from an exhaustive review — an idea's absence from this table means it was not found, not that it does not exist. 🟡
12. The elite split and the pretext: not a rival to the protocol but its application¶
The architect's objection to §11: "an elite split is more likely with 90 % popular support. And elites need a pretext for consolidating with the people and among themselves. A pretext that is explicit and cannot be openly condemned."
The assistant set the elite split up as a rival theory under which the protocol is secondary. That is an error, and here is where it lies.
A split fails not on dissatisfaction but on coordination¶
There is never a shortage of discontent inside an apparatus — that is not the scarce resource. What is scarce is the first move: nobody can move first, because nobody knows who would follow, and a lone defector is destroyed at no cost to the regime.
That is a coordination problem, solved not by persuasion but by common knowledge — the very mechanism 056d.7b rests on: what matters is not how many know but that each knows the others know (Chwe, Rational Ritual). What is needed is a public focal point around which people can converge without conspiring.
Hence mass support is not an alternative to a split but the input to it. The architect is literally right: with visible support a split becomes likelier, because its principal blocker is removed.
Why the pretext must be of this particular kind: four requirements¶
| Requirement | Why | Does the protocol supply it |
|---|---|---|
| Uncondemnability | for "democracy" or "against the war" a defector is destroyed as a traitor at no cost | yes: demanding a share of the subsoil is not treason but populism, and populism is survivable |
| Measurability | one cannot coordinate on a feeling, one can on a number | yes: the counter and the anonymous count produce a figure (§7) |
| Consolidation with the people rather than away from them | a split on a liberal platform leads away from the 90 %, not towards them | yes: "for the people" rather than against the authorities (§10) |
| Personal gain for the defector | "democracy" costs him rather than paying him | yes: he is a recipient too, as is the foreman (§8) |
The fourth row is one the assistant did not see at all. Every previous platform asked an official to give something up for the common good. The protocol pays him too: he is a citizen, the share is due to him, and no self-sacrifice is required. That removes from a potential defector precisely the argument that has held him so far.
And the defector's armour is the regime's own rhetoric¶
001c already contains the fork: "the dictator and the defector from the elite — kill him and make a martyr; leave him and he prepares the transition" (048d.6, 057c.4). The protocol adds a third side that was not there.
A defector saying "Russia is rich and a share is due to everyone" cannot be condemned without condemning one's own propaganda: the riches of the subsoil are an official thesis and a matter of state pride for decades (§7). A traitor is shot; a man repeating the television with a question at the end is shot with explanations, and there are none to give.
So the protocol supplies not only a pretext but armour, cut from the regime's own words.
What follows for §11¶
The "elite split" row is removed from the table of rivals. The correct formulation: the elite split is the mechanism, and the protocol supplies its missing part. The theory is right, and precisely for that reason mass work does not contradict it but serves it.
And the addressee of §4's four actions changes. They looked like work with the population. It is now clear that the calculator and the anonymous count produce exactly the public landmark an apparatus needs in order to coordinate: a figure that can be cited without anyone conspiring with anyone.
The degenerate case is unstable: the correction that closes the last hole¶
The assistant left the principal objection open: an apparatus will take the payment without the formula, and the mechanism has no defence. The architect closed it:
"A payment without a formula immediately raises the question: how did you calculate it, maybe I am being underpaid? And there will be a formula, and many of them will circulate among people, and the opposition will throw in its own, which will also be discussed."
Right, and it is the same device as everywhere in this chapter: a number without a rule itself provokes the question of the rule, and the question is asked by the recipient, in his own interest, with nobody organising it.
The payer's fork closes:
| Move | How it ends |
|---|---|
| Publish a formula | that is the audit of the rent upon itself (001c, the incumbent's row); thereafter the size has to be defended aloud |
| Do not publish | every recipient supplies his own guess, the guesses spread through the same private channels (§10), and refuting them requires publication — that is, the first move |
So "without a formula" is not a state but a pause. The absence of a rule holds exactly until the first payment; after it the counting starts, and everybody counts.
And the calculator's role changes for the third time. First it was a way to give a person a personal number (§5), then the trigger of a trap (§9). Now the main thing becomes visible: the calculator is a competing formula, permanently available. The moment the state names a sum, an independent computation of what it should have been stands beside it. Nothing extra is required — the instrument is already in place.
Add that the opposition will throw in its own formula, and the field turns into a comparison of rules conducted in numbers and in public. That is exactly the ground the protocol was built to fight on: an argument about a rule rather than about trust.
Hence the test sharpens: not "is there a formula" but "who holds the inputs"¶
The real degenerate case is not the absence of a formula but a formula whose inputs the payer controls: "five per cent of the surplus," where he defines the surplus. Such a construction looks like a rule and works as discretion.
So §9's test ("who sets the size?") takes a precise form: who controls the quantities the size is computed from? And from here it is visible why the formula is tied to median earnings from an independent source rather than to the volume of rent (§5). That was recorded as protection against manipulation. In fact it is what makes it unforgeable in a competition of formulas: a rival whose inputs are his own loses to anyone whose inputs are external — and loses publicly, because the comparison is computed on the same calculator.
Who will do the counting: the question answers itself¶
The assistant set it as a condition that "somebody has an interest in counting." The architect removed it in one sentence:
"Counting the fairness of money paid to me — people love doing that. And imagining: if they had given more, I would have bought…"
Right, and the condition turns out to be met by default: the interest in counting does not need creating, it would need suppressing. People count their own money constantly and unprompted; counting somebody else's requires a motive, counting one's own does not.
But the second half of the sentence matters more than the first, and it has a named mechanism. Value is felt relative to a reference point rather than in absolute terms (Kahneman and Tversky), and losses are felt more strongly than equal gains. Hence:
- without a formula ten thousand roubles is simply ten thousand;
- with a computation showing forty, the same ten thousand is felt as a loss of thirty.
So a payer handing out money in a world where a calculator exists buys not gratitude but a grievance whose size he does not set. And the more he pays, the more precisely a reference point forms — after which it cannot be given up.
"I would have bought…" is the form that reference point takes: not abstract arithmetic but a specific foregone purchase. It makes the gap personal and vivid — exactly the property §5 wanted when it required measuring in salaries rather than roubles.
Hence an addition to the calculator's specification (§5): show not only the sum and the gap but what the gap buys. Not "you were underpaid thirty thousand" but "that is a refrigerator" or "that is six months of utility bills." The unit should be a thing rather than a number.
Who the disappointment lands on: the assistant assigned blame wrongly¶
The assistant recorded the risk thus: if the number proves unattainable, the disappointment attaches to whoever named it — "they promised forty and gave ten, so everybody lies." The architect objected:
"You have forgotten again: the opposition is unkillable in the information field, and people support it. The authorities will come under heavy attack."
The correction is accepted, and the error was a substitution of terms. A computation is not a promise. §5 requires outright that the formula be open, the figures changeable and the source of every line given: the calculator says "here is the arithmetic, check it," not "we will pay you this much." Whoever pays is the one who promises.
Hence an asymmetry the analysis lacked:
| Who names a number | What he owes | |
|---|---|---|
| The opposition | computes and shows the computation | nothing: it does not pay and promised nothing |
| The authorities | pay and show no computation | an explanation, publicly and continuously |
Blame follows whoever cannot show his working. A payer without an open formula loses the argument not because he is wrong but because he is the one who owes an answer.
And "unkillable in the information field" needs refining, because §10 says otherwise. Opposition channels are weak as distributors — most people do not read them. But they are indestructible as a source: they publish from abroad, cannot be closed, and what is published stays available.
For an argument about a formula that is sufficient and nothing more is needed. Keeping a formula in circulation requires not mass reach but the impossibility of being silenced. After that others repeat it, and repeat it in the very transmission medium §10 describes as the principal one.
Meanwhile the authorities must hold the line continuously: the money has been paid, the question has been asked, and silence is itself an answer. It is a war of attrition the stronger side loses by construction rather than by strength.
What remains of the risk¶
One thing, and it is an argument about quality rather than blame.
If a number is named as a promise rather than as a computation — inflated for effect, with no arithmetic shown — it becomes a promise in fact, and then nothing written above protects it. So the discipline remains, on a different ground: a defensible number is needed not to avoid blame but to win the competition of formulas (above). A computation that cannot be checked loses to one that can, regardless of whose number is larger.
Weak point, and a serious one. No case is known in which a demand for a resource dividend produced an elite split — the mechanism is derived from general coordination theory rather than observed; it is one of the two things with no analogue at all (§14). "Ninety per cent support" is the architect's assumption. The reasoning about the instability of a payment without a formula rests on Alaska 2016–2018 (§13) — the nearest case, but not direct confirmation: there the payment is not tied to voting. The claim that the opposition cannot be silenced is untested: it holds today and is not guaranteed tomorrow. And one condition the architect did not remove remains: there must be something to count with — a calculator must exist and rent figures must be published. The first is within our power, the second is not. 🟡
13. Method: what replaces the caveat "no data exist"¶
The section stays here because this is where the rule was found; in general form, together with two others, it is recorded in 023b.
The architect's remark during work on this chapter:
"You have worn me out. We are reasoning about something that NEVER EXISTED, and yes, a pilot is best. BUT in forecasting we must minimise the error of our assumptions and look for confirmed indirect or adjacent entities. And you write me the same thing every time — that is inefficient."
The remark is just, and the failure is twofold. First, the absence of direct data about something that has not happened is a constant rather than a finding, and repeating it in every section is pointless: this chapter accumulated twelve variants of "not measured" and forty-three markers. Second, an excuse was substituted for work: the nearest confirmed analogue lay in the repository itself and was not opened.
The rule replacing this: where direct verification is unavailable, no caveat is written until the nearest observable analogue has been found and how far it carries has been stated. A caveat is admissible where no analogue exists at all — and is then written once per chapter rather than at every claim.
A table of adjacent entities for this chapter's assumptions¶
The third field has been rewritten in the form of 23b.1: what transfers, which way the gap shifts it, how much remains. In the first edition it was binary ("does not carry"), which read as a cut-off on the transfer, whereas a difference in conditions is a discount.
| The chapter's assumption | Nearest confirmed analogue | What transfers and how much remains |
|---|---|---|
| Money moves the apathetic towards participation | James et al. (2025): +10 % in the size of Alaska's dividend (~$190) → +1.4 pp of turnout, with the effect among the young and less well-off; Loeffler (2022) — the dividend raises turnout (014) | Transfers: the relationship, its sign, and the target group — the very one §5 concerns. The gap runs both ways: Alaska is a democracy where turnout is not suppressed, so there is less headroom → the analogue understates; but the payment there is not tied to voting whereas ours is → it overstates. Remains: largely, with the sign of the net shift unknown |
| People turn up for money | Alaska 2024: 666,213 applications, 624,489 payments against a population of ~735,000 — about 85 % of residents receive it, children included (014) | Transfers: the existence of the phenomenon and its scale — almost everyone turns up for money. The gap: the number of eligible non-applicants is unpublished, so no take-up rate can be derived and the 85 % is itself a lower bound. Remains: largely, as a lower bound |
| A payment without a formula is unstable and turns into an argument about the rule | Alaska 2016–2018 (014.14.3b): a discretionary halving → a lawsuit, collapsing approval, the governor leaving the race, a rival winning on the full formula | Transfers: the whole chain — a cut, a lawsuit, political consequences. The gap: the causation is documented as press interpretation rather than measured, and the immediate occasion of the withdrawal was the lieutenant governor's resignation. Remains: largely, with a reservation on causation |
| The formula must be protected more strongly than anything else (004.4.7b) | Wielechowski v. State (2017): a statutory formula does not bind legislature or governor | Transfers in full: this is a judicial ruling about a property of the construction, not about a country. No substantive gap — the lesson applies in any jurisdiction with annual appropriation. Remains: almost entirely |
| Envy works for distribution rather than against it | Foster (1965) + Chinoy et al., AER (2026): zero-sum thinking predicts both support for redistribution and harshness towards outsiders (056d.5b) | Transfers: the structure of the relation and both correlations at once — they match both of our objections. The gap: no measurement for Russia, so what is unknown is the magnitude, not the direction. Remains: direction confidently, magnitude not at all |
| Common knowledge changes an apparatus's behaviour | East Germany, May 1989 (056d.7b); Chwe's coordination theory | Transfers: the existence of the phenomenon — accumulated common knowledge changes an apparatus's behaviour. The gap is large: a single case, no money question in it, and §1 supplies a direct observation against immediate action. Remains: the existence of the phenomenon only, and with contrary evidence beside it |
What remains without an analogue — and that is the chapter's entire caveat¶
An analogue is missing for two things and only two: a payment tied to the act of voting (never done anywhere), and an elite split caused by a dividend demand (no case exists). Everything else in the chapter rests either on a measurement or on an observable case.
The earlier repeated caveats are accordingly withdrawn: where "not measured" stood, a row of this table now stands.
13b. Consolidation on an external figure: Aleksashenko's 75% and a dividend of 40–50 thousand ₽¶
(Added 2026-09-29.)
The starting figure is not the architect's assumption. The economist Sergei Aleksashenko, on TV Rain on 27 September 2026 ("На пороге бедности. Алексашенко — о том, что будет с экономикой в ближайший год", timecode 17:24–17:57), reported the result of a survey conducted in Russia by his team in two rounds with the same outcome: three quarters of Russians (75%) believe the war is damaging the Russian economy. The question was indirect — not "for or against the war" but "does it improve or worsen the state of the economy" — so the distortion from fear is smaller than usual. Under the rule of §13 this is a lower bound: the distortion in Russian polls runs downward.
Aleksashenko himself concludes that people will not be able to show their opinion: "I would not expect Russians to take to the streets." This is exactly what the protocol answers: it offers two lawful moves, and both express disagreement.
The architect's reading: all 75% are against the authorities; the only difference is how they express it — with a vote (A) or with money (B). Whoever takes the money does not vote for the authorities — he stops being raw material for herded turnout.
The dividend. The sum follows the architect's rule of 29.09.2026: the dividend can be any amount; the single measure is that it must be important and desirable for 40–50% of the population. For Russia that is 40–50 thousand ₽ per cycle (≈0.6–0.75 of the median monthly wage of 65,307 ₽, §5). Cost: with 55–66 million recipients, 2.2–3.3 trn ₽ per cycle, i.e. 440–660 bn ₽ a year, about 8–12% of the 2025 deficit (5,644 bn ₽, §5).
Parameters of the calculation.
| Parameter | Value | Source |
|---|---|---|
| Against the authorities | ≥ 75% | Aleksashenko, 27.09.2026, 17:24 (lower bound) |
| Ideological core of the regime | 10–15% of the electorate, stays in A at any sum | 056e |
| The rest of the 25% | 10–15%, take B | 056e: administratively dependent and apolitical |
| Share of B at D = 40–50 thousand ₽ | 40–50% of the electorate | rule "desirable for 40–50%" |
Consolidation No. 1 — support for the protocol itself in a referendum.
| Group | What it gets | Position |
|---|---|---|
| Against the authorities, vote (A) | vote weight ×1.5–2 | for |
| Against the authorities, take the money (B) | 40–50 thousand ₽ | for |
| Loyal out of habit, take the money | 40–50 thousand ₽ | for |
| Core of the regime (10–15%) | loses its result | against |
Support for the protocol: 85–90%. This independently matches the estimate in 013c.5 (against: 10–16%).
Consolidation No. 2 — the election result after the protocol. If 40–50% of the electorate goes to B, then of the 25% everyone except the core (10–15%) goes there, and the remaining 25–40% of B are opponents of the authorities who chose the money.
| Share of the electorate | |
|---|---|
| B — money | 40–50% |
| A — against the authorities | 35–50% |
| A — for the authorities (core) | 10–15% |
Against the authorities: 70–83% of votes cast; for the authorities: 17–30% (today the authorities get 70–80%). The weight of one vote: ×1.5–2.2.
Safety margin. For the authorities to lose the majority among those voting, it is enough for 13–20% of the 75% to stay in A — that is, every fifth to seventh opponent of the authorities. The rest can take the money; the result does not change.
| Today | After the protocol | |
|---|---|---|
| 75% against the authorities | dissolved in herded turnout, the vote weighs nothing | vote with weight ×1.5–2.2 or take the money lawfully |
| Administratively dependent | brought in by the boss, they deliver the regime's result | take B; the machine has no one to bring |
| The authorities' result | 70–80% | rests on the core: 17–30% of votes |
All 75% are against the authorities and express it differently: some with a vote, others with money. 85–90% of citizens would support the protocol. At the election the authorities would get 17–30% of the votes. Opponents who choose the money do not weaken themselves — they strip the machine of its turnout.
A check on the margin: what if the 75% is halved. The survey measured an assessment of damage to the economy, not an intention to vote. The architect's objection: "even if you divide by 2, the margin is enough." The check confirms it. Suppose only 37.5% are against the authorities and the other half are neutral. Then the neutrals (37.5%) together with the non-core part of the loyal (10–15%) make 47.5–52.5% — enough on their own to fill B (40–50%) without touching the opponents of the authorities. Up to 37.5% against the authorities remain in A; for the authorities — the core of 10–15% plus at most 12.5% of the remaining neutrals. Even in the worst case, 58–79% of votes cast are against the authorities. And to outvote the core alone, it is enough for every third of these 37.5% to stay in A.
The sum is not the people's question, and an error in it is cheaper than dictatorship. 40–50 thousand ₽ diverges from the formula of chapter 01 (K = 1% gives about 11 thousand ₽, §5). The architect's answer: "the size of the payments is not decided by the people; if the people err but bring down the dictator, then there will be a democracy discussing a reduction of the payments — and which is better?" Compared with the present state of affairs:
| Today | After the protocol with an "overstated" sum | |
|---|---|---|
| Cost | the war and its expenses; a deficit of 5,644 bn ₽ in 2025 | 440–660 bn ₽ a year — about one year of state support for agriculture (665 bn ₽, §5) |
| Who decides about money | authorities who cannot be replaced | authorities who can be replaced, plus a referendum on the formula (001b.1) |
| If the sum turns out to be a mistake | nothing can be corrected | the mistake is discussed and corrected openly, by rule rather than by decree |
An overstated sum is a mistake a democracy can correct. Dictatorship is a mistake that cannot be corrected. Hence for the transition the sum is chosen by a single measure — desirable for 40–50% of the population — and its exact size after the transition becomes the subject of an ordinary open dispute. Synchronisation debt: the charter of 049 and chapter 01 still hold K = 1%; the rule "desirable for 40–50%" has not yet been carried over into them.
13c. Shulman, 30.09.2026: no results — six tasks and one construction¶
The occasion: the episode "Why are there no election results?" (30.09.2026, 8 minutes). The architect: "once again Shulman gifts AB-EXIT tasks she herself does not even know how to solve." Her figures are given as she states them; the protocol answers each with the same construction — double entry (048f) and three channels (048i).
| What Shulman says | The task | How the protocol closes it |
|---|---|---|
| There are no results, but the mandates are distributed and the Duma is convening; for precincts covering ~6 million voters nothing is published: Sakha minus 10 %, Kursk region — 84 %, Leningrad region — zero | The count is the commission's private business: not published, and there is nothing | The treasury has its own record, independent of the commission: the number who took the sum per precinct; the turnout ceiling = the roll minus those who exited (048f §3.2). A precinct without results is a precinct without a balance (048i §5b), an anomaly from day one. The money of those who chose nothing waits for them all cycle (048k §3): every uncounted precinct has people who will come for the payment and ask |
| The targets were changed three times: from the administration — then higher, when the president declared the election a plebiscite — then 60,000 ballots to drag SR over 5 %; "like a bad accountant who stole the money and now has to make the spreadsheet add up" | The count is fitted to a target, and the target floats | There is nothing to fit: the number of B is recorded in the treasury before polling day, the number of A is capped by it from above, the tokens of the open channel give a lower bound; rewriting the result after publication means diverging from the two other records (§1, the Katz amendment) |
| Four "imaginary regions": zero published results, but "non-existent souls" form the parliament | Dead souls vote | Every record closes with one of three outcomes: a vote, the sum, or silence (048i §5b). A dead soul cannot take the sum to an account on a document — it stays "empty", and the empty are the zone of special control (001c §3); a sum paid to a non-existent person is embezzlement visible in the treasury |
| Every vote is worth 152 roubles of budget money a year; United Russia will get another 1.5 billion for what was drawn, LDPR plus a half, New People from 455 to almost 800 million; "ten old women — a tenner" | The state itself pays parties for drawn votes: the incentive to draw is built into the law | The protocol reverses the addressee: the budget pays the citizen for exit, not the party for a vote (055c §6.6, 029.4b). A drawn vote stops bringing money to whoever draws it |
| Vologda: drawn for the KPRF (31 % on the lists) to set up the governor; Samara: votes taken from Matveev and added to the approved communist | The count is a weapon in intra-elite struggle | The weapon disappears with the monopoly on the count: three records with three owners, damage to one visible from the other two (048f §4, 048i) |
| "Impunity has a reverse side": external control, transparency and accountability to the taxpaying population were a protection for the bureaucracy itself; without them "the town falls asleep, the mafia wakes up", and "there is nobody to go to" — no instances of appeal | An external constraint is needed that the power will accept itself | The protocol is a constraint the bureaucracy will accept, because it pays it too: the policeman wants the dividend (048d §7); six disputes decided within days (048j) are the instance of appeal that, by her words, does not exist |
The main thing in the episode is the last point, and she put it more exactly than we did: the absence of control is most dangerous for those who removed it. This is an argument for the protocol from the mouth of the bureaucracy, not the opposition. Her conclusion "there is nobody to go to" receives an answer: there is nowhere one needs to go — everyone has on his account a record that cannot be rewritten, and a person who will come for it.
13d. Shulman on TV Rain, 01.10.2026: remote e-voting — a channel built for coercion that turned out the most honest¶
The occasion: the interview with Yu. Taratuta "We are in a space of escalation" (TV Rain, 01.10.2026, 39 minutes; transcribed locally, the video had no subtitles). The architect: "what is interesting here for us, especially on remote e-voting". Her figures are as she states them.
What she says about remote e-voting (DEG).
- Two electronic systems that do not match each other: the Moscow one and the federal one; she has not seen the Moscow results.
- DEG is the instrument for coercing public-sector employees: vote from the phone on Friday morning at the workplace, under supervision, identically from Sakhalin to Kaliningrad — an instruction from the centre. Used in 32 regions, mostly not in the "electoral sultanates": where paper cannot draw the result — distributed "rather cleverly".
- The technical operator of the federal system is the FSB: "a black box, turn it any way you like, nobody will stop you"; Moscow 2021 — "the electrons out-voted the people" in eight districts of fifteen.
- The first federal DEG results at night: United Russia 49 % — below "the psychologically important level", New People second, the KPRF third, LDPR and SR below 5 %. By morning United Russia had 50 %, second place unchanged. The contours match the statisticians' estimate separating the vote from "the white noise of falsification": turnout 40 %, United Russia 35 %. "The Russian collective public-sector employee did not demonstrate what was wanted of him" — knowing the boss sees his click, he did not click United Russia.
What follows for the protocol.
| Her observation | For the protocol |
|---|---|
| The channel built for coercion turned out closest to the truth, because it was published before it could be rewritten: the night's 49 % became the morning's 50 | Post-publication rewriting (§1) caught in the figures. Publish the counter by precinct during the window, machine-readable, with no right of suspension (004 §4.7b): what is published at night cannot be rewritten unnoticed |
| Coercion buys the click, not the choice: brought in, voted against | Confirmation of 048h §6: bringing in is free, the choice is not. After the protocol whoever is brought in under button A loses the sum — coercion for the first time costs him money he can see; pressure on the choice is a criminal offence (048k §4b) |
| One operator — one black box | The protocol's open channel is electronic too, and this is a warning: it cannot be one agency's app. The token is verified by the voter himself and by the candidate's server (048i §2, §4), channel B goes through the banks — a second owner of the record by construction. Three records with three owners against one with the FSB |
| Two systems that do not match | An argument for a single treasury register: the sum paid once to one person — one record for the whole country (048f §3.2) |
The main conclusion — the architect's: the power itself trained people in the instrument. "What matters for us is that DEG is introduced by the authorities, and they themselves train people to use it: our tokens are already proven — people can and will use them, for AB-EXIT and for choosing the money. Thank you, authorities. And adults and the computer-illiterate use it too, if they are asked very insistently." The main objection to the token and the app — "the old will not cope, the village will not understand" — is removed by someone else's hands: the state opened accounts for everyone, taught people to vote from the phone on Friday morning, and taught precisely those considered incapable. The button "money or vote" is simpler than a DEG ballot: there a choice among parties, here between two buttons. The usability pilot has been run on millions.
And second: "that DEG produced an almost protest vote despite all the difficulty means that if a poor rural public-sector employee had the choice of taking 30,000 roubles, his bosses on Monday would hardly have forced him to vote for free — and you cannot fire all the public-sector employees. In short, the power is already struggling, and with AB-EXIT it is simply brutally hard." Coercion breaks on the sum: the one brought in takes the money, and the boss has nothing to contest it with — "I am not against the power, I need the payment more" (057b §4); forcing a person to lose a month's wage is harder than making him press a button for free (048h §6). The main instrument of bringing people in turns into a paid exit.
Third — Russia as a test range for the objection "the country will collapse". The architect: "if the opponents of AB-EXIT say the country will collapse if a large share of the budget goes to payments — Russia did not collapse from a gigantic share on war, that is, from money thrown away into nowhere at all." The figures: defence in the 2025 federal budget — 13.5 trillion roubles, 32.5 % of expenditure (SIPRI); in 2026 defence and "national security" together — 16.8 trillion, 38 % of the budget; the plan for 2027 — 17 trillion (Reuters). The dividend by the yardstick of §13b — 40–50 thousand to 110 million voters with half exiting — is 2.2–2.8 trillion, that is, 5–7 % of the federal budget, about a sixth of the military-and-security line (the calculation in §5: 2.2 % of one revenue line). The difference is not only in size: military money goes into metal and abroad, the dividend stays in the country and is spent at the nearest shop. A country that has withstood 38 % on war and guards for four years will withstand 6 % on its own citizens.
The rest of the episode that will be of use.
- "The top sees the folder": the leadership does not know the real figures, the picture is clearer to those closer to the ground; those who report and the one who listens sustain the illusion mutually. This is 029.10 "the captured instrument" live: a poll and a protocol the power can draw for itself, the number who took the sum it cannot; the thermostat is honest for the one at the top as well (048d §4).
- Governors compete over "whose bag of votes is heavier", and mobilise from the same obedient poor territories where they draw: "much noise, little wool" where United Russia gets little. With a turnout ceiling set by the register of those who exited, the bag cannot be inflated (048f).
- No results for a week: "the accountant stole the money and now has to make the spreadsheet add up" — last-minute rewriting for 60,000 ballots for SR; double entry makes it impossible (§13c).
- "Smart voting is the most effective of all the ineffective instruments": her own estimate of the ceiling of the present repertoire (056d §7).
13e. The admission at 27:55: "effectiveness cannot be spoken of" — and the move missing from her repertoire¶
From the same interview (TV Rain, 01.10.2026, 27:55–28:16): "at authoritarian elections one cannot speak of effectiveness, the predetermined result will be presented… we cannot speak of effectiveness… we call nobody to anything." The architect: "she admits the complete failure of any policy and action of the opposition — no election can be won or obstructed. But the protocol would really obstruct even the present elections: people en masse would ask the power why they cannot have the money — millions of questions."
Why her repertoire is helpless. Smart voting, boycott, spoiling the ballot, observation — all are moves inside the count, and the count belongs to the power (048f §4e). A move whose outcome the commission records will be recorded as required; that is what "the predetermined result will be presented" means. The opposition looks for how to win a game in which the opponent keeps the score.
Why the question about money does not belong to the count. It cannot be drawn, recorded as "50 % for", or declared foreign-agent: it has no candidate, no party and no "against" — only a sum the power has already proved it can pay (§13c: 152 roubles per vote to the parties; 029.4b), and a refusal with one meaning: "we need your vote for free". Any answer exposes: "there is no money" — against 16.8 trillion for war and security (§13d); "this is bribery" — and the 152 roubles to the parties?; "you vote anyway" — yes, under coercion.
Two properties none of the moves on her list has:
- The question does not require winning an election. It exists before them and apart from them; it is the only move that does not pass through the ballot box, and hence the only one that cannot be predetermined.
- It is the kitchen, not the square (039). Asking "where are my 30,000" is not politics; there is no risk. Her own evidence: the public-sector employee under the boss's eye did not click United Russia (§13d); whoever does that at a risk will ask about money without one.
A correction to "millions of questions". They will not arise by themselves: the question must be sown, a carrier is needed even for the kitchen (041). But he needs nothing except the word "dividend", which everyone likes (033c §9d).
Weak point of the move. The first version named the capture of the word: the power would pay a "dividend" for turnout. Withdrawn on the architect's correction (§13j): to pay on demand is to bend and is the power's most expensive move, and payment for a vote and payment for exit are opposite signs of the thermostat (029.4b). The real weak point is silence: neither paying nor refusing (§13j).
13f. The Bonya case, April 2026: the channel of §13e has already worked — with the wrong question¶
Facts. On 14.04.2026 the blogger Viktoria Bonya (Monaco, 16 years outside Russia; Forbes list of the richest bloggers) posted a ~19-minute address to Putin on Instagram, blocked in Russia: over 20 million views and a million likes within a day. On 15.04 Peskov confirmed that the Kremlin had watched the video and "work is under way"; Solovyov attacked; on 16.04 she thanked Putin and Peskov in tears and told TV Rain and the BBC "I am not with you". Five topics: the flood in Dagestan, fuel oil in Anapa, the seizure of cattle near Novosibirsk, the blocking of Telegram and Instagram, the Chekalin case; plus a request to create a "direct platform" for complaints to the president. The frame: "the people fear you", "the governors lie", "the power serves the people, not the reverse". The architect: "if this case is laid over the protocol — what do we get? She spoke of things the people do not even much need; what if her address had been about the protocol".
The channel of §13e is proved empirically. A non-opposition carrier, zero organisation, a blocked network — and an obligatory Kremlin answer within a day. A question that does not pass through the ballot box the power is forced to answer; this is an April fact, not a hypothesis.
How her questions were closed — with words. "Work is under way" cost nothing, and within a week the hype was over: the topics were local, a flood in Dagestan concerns Dagestan. The question "pay for the vote, you take it anyway" cannot be closed with words: the answer is either 2.2–2.8 trillion (§13d) or a refusal — the first public pricing of the vote by the Kremlin. "Work is under way" becomes the meme "when are the 30,000?" and is renewed every polling day, because the money has not come. One question instead of five, and one for everybody, not for a region.
Her main thesis is our instrument. "The governors lie, the president does not know" — this is "the folder" (§13d) and the captured instrument (029.10). She asks for a "direct platform", that is, voice upwards (036c, Hirschman); the thermostat is the same instrument but honest for the top as well: the number who took the sum can be drawn neither by a governor nor by the e-voting operator (048d §4). She asked for what the protocol already has.
Her million likes are the protocol's audience. The apolitical, without a kitchen, those the populists hold; the word "dividend" was made for them (033c §9d), and even a blocked network reaches them.
What we would have got. An address about payment for the vote can be called neither opposition (she herself: "I am not with you") nor foreign-agent; one video sows the millions of questions of §13e — a carrier of 041 at zero cost.
"And this is only one blogger of thousands" (the architect). This changes not the scale but the risk. One carrier is bought with attention or switched off; a thousand are not. Bonya had five topics; a thousand bloggers have a thousand topics, each closed with words. With the protocol the thousand have one question with one word (033c §9d): the answers add up to one count, and "work is under way" cannot be said a thousand times. No coordination is needed — the blogger has an incentive of his own: 20 million views are money, the question about the 30,000 is monetisable content. The large group gets a focal question without organisation (015c §7b).
Weak point — the carrier. She gave thanks in tears for having been watched: a carrier satisfied by attention is bought with attention. The protocol needs from the carrier not loyalty but the question, and the question outlives the carrier: it cannot be "watched and closed". A one-off payment "following the address" is not a risk: it is the power bending, not its move (§13j).
13g. The split of the pro-regime camp over Bonya: the dividing line is ours, the route runs through the loyalists¶
The architect's question: "did the pro-regime Z-bloggers, who are no opposition at all, support Bonya or keep silent — and how does this help the spread of the protocol?" The facts of the reaction (April 2026; she has 13 million subscribers, the video ended at 30 million views and 1.6 million likes).
They attacked the person, not the content. Podolyaka (2.8 million): "an enemy to all of us, like Navalny", "part of a plan to rock Russia"; Kirill Fyodorov: "rituals with banned substances, fled to Monaco"; Life — "lives in Monaco"; Tsargrad — a segment "Glamour junkie against Putin"; Solovyov — "a worn-out prostitute" and a request to the Ministry of Justice for foreign-agent status; Milonov — "escort". The Kremlin bots doubled their norm: 16 comments per post instead of 8. Not one disputed any of the five facts.
Support came from the same pro-war milieu. Kashevarova: "she said what we have been writing for a long time"; Kolyasnikov, host of Solovyov Live: "the first fruits of the idiotic blockings"; Dva Mayora: "her content interests people more than the war". The insults produced a second wave larger than the first — the "women's revolt", a flashmob and a class action.
What follows for the spread.
- The dividing line is the protocol's line. Since 2022 the Z-milieu lives in the frame "the president is lied to, the middle steals"; Bonya landed in it literally. The protocol's question lands there too: the governors' "bags of votes" (§13d) are the middle, the thermostat is an honest number for the top (048d §4). In Z-language the protocol sounds like "give the top the truth no governor can draw".
- The route runs through the loyalists, not the opposition. To TV Rain she said "I am not with you", and for a carrier that is right: the audience of opposition media is a few per cent, that of Z-channels and lifestyle bloggers tens of millions. A host of Solovyov Live carried her thesis: the power's own infrastructure carries the question if it is not against the power (056d §7b, 035).
- The power has two answers, and neither works against a sum. "Work is under way" closes a complaint but not "where are my 30,000". "An escort from Monaco" refutes neither Dagestan nor the sum, and produces a second wave: an attack on a non-opposition carrier hits the attacker. Hence the choice of carrier (041): one off whom ad hominem bounces — a mother of many, a soldier's wife, a rural public-sector employee; "foreign agent" does not stick to them.
- The bots are a counter. The doubling of the comment norm is a measurable sign of a hit and a free metric of spread.
- "Interests people more than the war" — a demand signal from the war audience: a domestic money question outranks the war. The dividend of 5–7 % (§13d) is to be presented not as "instead of the war" but as "the money exists". Per §13l this is a matter of speed, not of admission: anyone may raise the protocol.
The architect's correction to the first version of this section. I wrote "the vote is the top's sacred thing, and 'rocking' will be heard more often". The architect: "no: the people's bloggers, the ones not bought, will at once raise that the parties are paid and the people get no money — what has the sanctity of the vote to do with it when there is nothing to eat". Correct: "the sanctity of the vote" is the vocabulary of the count, and the protocol's question does not belong to the count. A people's blogger will not say "the elections are fake"; he will say: "152 roubles for every vote are paid to the parties by law (§13c), and to the people — nothing". It is the same anti-bureaucratic line of the Z-milieu, only the addressee is not the governor but the party till, and "rocking" cannot close it: the figure comes from the law on parties, not from opposition media.
Weak point. The split over Bonya rested on her questions costing the Kremlin nothing: supporting her was free even for Kolyasnikov. The money question costs 2.2–2.8 trillion, and some of the supporters will step back when they grasp the price; how many, only the first carrier will show.
13h. The populist's weapon, but smarter: an arsenal of comparisons for the people¶
The architect, 01.10.2026: "here one can fight the populists with their own weapon, but smarter: a deputy is paid 150,000 roubles a month for lunches, and I cannot be given 30,000 once in four years — and similar beautiful but not quite honest comparisons for the people".
The rule. All the figures are honest; only the juxtaposition is not: a month against four years, a lunch against the dividend. That is the "smarter": the populist's figures lie, ours lie only in scale, and checking every figure works for us. The second difference: the populist promises out of a budget he will distribute himself; the protocol's question asks not for a handout but for a rule (048k §3, "the right, not the methods"). All the comparisons have one form: the vote is already paid for, only not to the one who votes (029.4b).
| Comparison | Figures (2025–2026) |
|---|---|
| "A deputy gets my four-year dividend in one working day" | a State Duma deputy's pay ~520,000 roubles a month ≈ 23,600 per working day; the dividend of 30,000 once in four years = 625 a month |
| "For my vote the parties are paid, I am not" | 152 roubles per vote a year = 608 per cycle; United Russia in 2025 — 4.3 billion of state support (§13c) |
| "Counting my vote costs 212 roubles, the vote itself — zero" | 23.28 billion roubles for the 2026 Duma election / ~110 million voters |
| "625 roubles a month — and on this 'work is under way'" | Peskov's reply to Bonya's address (§13f) |
| "The money exists: the dividend is two months of the line 'defence and security'" | 16.8 trillion a year = 1.4 trillion a month; the dividend 2.2–2.8 trillion (§13d) |
The last row follows the rule of §13g: "the money exists", not "instead of the war". The word is "dividend" (033c §9d); the addressee is the middle and the party till, not the top (§13g). The arsenal is replenished by country: each has its own deputy, its own party payment and its own bill for the count.
Division of labour (the architect: "the opposition in exile will find such figures — fireworks for city day and other showy spending of the power — you need not bother"). Correct: showy spending has been collected for twenty years, the arsenal is ready and the habit is there. The protocol adds one thing to other people's figures — the denominator. "Fireworks for 500 million" by itself is indignation; "fireworks = the dividend of 17,000 people" is a bill. Without the sum a spending figure has no unit of measure; with it every outlay of the power converts automatically into "how many voters were left without the dividend". That is the protocol's contribution to others' work: not to find the figures but to give them a unit.
13i. The question instead of smart voting, and a sign on the ballot¶
The architect, 01.10.2026: "how real is all this for the opposition to implement as a replacement for smart voting, and how much better is the effect; one can also add a symbol on the ballot — but that is a secret weapon, to be brought out right before the election".
Who carries it — nobody. The opposition in exile as carrier is a poor option (§13g): its audience is the same 5–10 %, and the label turns the loyalist channels away; its role is §13h, the figures and the first seeding. I had written "a carrier from the loyalist milieu"; the architect corrected: that is still one carrier, and one gets removed. Yabloko was struck off the ballot on 10.08.2026 on Rodina's suit — for "May There Always Be Sunshine" in its campaigning, supporters' posts on blocked networks and "foreign financing": a structure with an address was removed. "Every Russian has his own bubble, and all the small bloggers about seedlings and cats will, for the hype, ask their audience too: what do you think about money at elections; bloggers already scan the whole internet professionally for a hype topic." The question is not carried — it is found, as Bonya was found (§13f): not distribution but self-distribution of a topic scanned by thousands without an address, surfacing in every bubble. There is nobody to remove. A refinement from §13l: "nobody" means "anybody" — the first carrier's label affects the speed, not the sign of the result.
Where it happens and what the opposition does. The architect: "this is self-distribution inside the power's core electorate; the opposition need only find the painful topics and the comparisons with the protocol's payments and explain their full, hundred-per-cent legality". Three tasks for the opposition and no public role: the painful topics (§13h, division of labour), the comparisons with the sum (the denominator) and the legal note — asking, comparing, writing on the ballot are legal; the question campaigns for nobody and against nobody.
The "on the stumps" precedent. The architect: "the referendum on the stumps proves that referendums can and even should be held". Stronger than it seems: the constitutional law on referendums forbids putting budget questions to one, and in 2020 the power went round its own law — invented the format "all-Russian vote" within a month, seven days, ballot boxes on stumps and in car boots, and Moscow paid for turnout with the "Million Prizes" lottery. The power itself proved two points of the protocol: a voting format can be created from scratch, and participation can be paid for (029.4b). Citing its precedent is legal by construction.
How the question beats smart voting.
| Smart voting | The question "where are my 30,000" | |
|---|---|---|
| Coordination | a list of candidates and an app that gets blocked | none needed: one word (033c §9d) |
| Risk to the participant | a link to an "extremist" structure | none: not politics, not "against" (039) |
| Legitimation | participation recognises the result | recognises nothing |
| Lifespan | dies on polling day | renewed every cycle while the money is absent (§13f) |
| Where it is measured | inside the power's count — "effectiveness cannot be spoken of" (§13e) | outside: knowledge of the sum in polls, the Kremlin's answer, bot activity (§13g) |
Where it is no better: the question also does not change the result of the current election and yields no mandates. It changes the agenda of the next cycle and takes the story of the elections away from the power; the effect is measured in people who know the sum, not in seats.
In per cent. The architect: "Katz spent tens of minutes and several videos on a million-strong audience for smart voting and got nothing, in Shulman's words". Katz's figures are an estimate from open data (channel ~2.5 million, videos 1–2 million views each); the question's figures come from the single Bonya precedent (§13f).
| Smart voting, 2026 | The question "where are my 30,000" | |
|---|---|---|
| Input | several videos, tens of minutes of flagship airtime | the same |
| Reach | unique viewers ~3–5 million = 3–4 % of voters (110 million) | Bonya: 30 million views from one video = up to 27 %; net of repeats and abroad 15–20 %; target 40–50 % — the protocol's yardstick |
| Chain to the result | saw → remembered the candidate of one's district → came → voted → was counted: five steps, the last done by the power | saw → knows the sum: one step, and it is already the result |
| Conversion to a measurable effect | 0 % — "effectiveness cannot be spoken of" (§13e); even at full conversion 3 % of voters ≈ 5 % of turnout, spread over 225 districts and rewritten (§13c: SR +60,000) | ≈ 100 % of reach: knowing the sum is the effect; plus an obligatory Kremlin answer (1 of 1 in the precedent) and a doubling of bots (§13g) |
| Residual value on 21 September | 0 — the videos are dead | full — the question carries over to the next cycle |
| Side effect | legitimation by participation; a label on everything touched (§10) | none |
In one line: smart voting bought 3–4 % of reach and 0 % of result, because the commission wrote the result; the question at the same price gives reach several times larger, and the result equals the reach, because the power is not allowed to count it. This agrees with the caveat of §5 ("a comparison of design, not of outcomes"): on 20 September the protocol gives no instruction, the question does — and it is carried out without adoption.
A sign on the ballot. The sign "30,000" turns the present ballot into a two-button one without any law: the person declares exit himself. Under Federal Law 67 a ballot with one mark in a square is valid whatever is written on it; without a mark it is invalid. Hence two modes: "sign + vote" — seen by the commission and observers, not in the result; "sign only" — lands in the line "invalid ballots" of every precinct's protocol, a published line, normally 1–3 %, which the power has never needed to draw. A jump to 10 % is visible across all precincts and cannot be attributed to the opposition: no candidate, no "against".
Two effects beyond the line: hundreds of thousands of commission members — the public-sector employees of §13d — hold the sign in their hands all day; photo-recording by observers makes the "invalid" line the first that is hard to rewrite.
Why right before the election. The question is sown for a year; the sign is announced in the last week — too late to brief 90,000 commissions or change the law. The sign explains itself if the sum is already known.
Weak point. The "invalid" line is inside the count (§13e) and can be understated; so the sign is a demonstration to commissions and observers, and the measure stays outside. The analogues — the "accordion" in Belarus 2020 and "Noon" in 2024 — the power ignored; the sign has what they lacked: a number in the protocol. 🟡
13j. The contract between people and power: the protocol does not break it, it prices it¶
The architect, 01.10.2026: "the contract between the people and the power — 'you stay out of politics, and we do not stop you living well'. Does the protocol break this social contract unnoticed? The people go into politics in order to live well".
The contract already exists and is already paid. "Loyalty in exchange for welfare" has been described since the 2000s; in 2014 the price was replaced by Crimea, in 2022 the state began openly paying for the most political participation there is — the service contract. The premise "politics is not paid for" has been dropped by the power itself. The protocol says it aloud: "you stay out of politics" — button B, "you live well" — D. The same contract with a signature under every clause.
What changes — who sets the price. Today the contract is a monopsony: one buyer of loyalty, the price set unilaterally and by mood (indexation, 10,000 to pensioners before the 2021 election, "Million Prizes" in 2020). The protocol moves the price to the median and makes it a rule (048k §3). The power keeps everything it wanted — the mass that stays out; it loses one thing — the right to decide what that is worth. That is the unnoticed break: the contract remains, and bargaining appears where there never was any.
Where it tears. The contract rested on politics and the good life being alternatives. The question "where are my 30,000" glues them together: a demand phrased as performance of the contract — "we stay out, pay us for staying out". It cannot be punished by the power's own definition (it is not politics, §13e); to refuse is to admit the contract is one-sided. The power is in the position of an employer to whom the worker has brought a bill for the first time.
Hirschman (036c). The contract bought loyalty — a state without price or term. The protocol replaces it with exit — an act with a price and a term of one cycle. Those who take the money are honest performers of the contract; those who remain have left it for politics, and their weight is paid for by the former (033c §9b). The population divides exactly along the contract's line, both sides get what they want, and the power loses the ability to count both sides as its own.
A handout is the power's most expensive move, not a cheap one. In the first version I named a one-off pre-election payment as the weak point. The architect: "you look at it wrongly: 30,000 is huge money for the power, and above all it is the power bending before the people, plain and vivid; a demonstration of weakness, not of legality; humiliating for the people and for the power alike. A lawful payment under a referendum everyone supported is not humiliating for the people, it binds the power, and the people owe it nothing for that. People will very much like obliging the power by law to pay them money — and a lot". Three consequences.
- A handout on demand is a concession under pressure. The 10,000 to pensioners in 2021 the power gave of itself and at its own time — a favour. To pay after the question "where are my 30,000" is to bend; the louder the question, the less room to pay without showing weakness.
- A gift creates a debt, a right does not. A handout is a gift: received — owed in gratitude, that is, in a vote; this is the mechanics of the old contract. A payment under law is the power's obligation, not its favour; it is humiliating to take alms, not humiliating to receive what is due.
- The pleasure of obliging. People like not only the sum but that the power is obliged — by a law they passed themselves. The only case in which the citizen is the state's creditor, not its petitioner.
A fork with no good move: to pay by handout is to bend; to refuse is to admit the vote is taken for nothing; to pay under law is to adopt the protocol.
Weak point. A fourth move remains — silence: neither pay nor refuse, while the question has to live without an event; the Bonya hype lasted a week. The answer is the single voting day every September: the occasion is renewed yearly, not once in five years.
Exit as an alibi: not a handout but a lawful "I do not take part". The architect: "the people whom the power forces to vote for it against their will very much need precisely a lawful exit from voting, not a handout before the election: they do not want to vote under coercion and cannot in conscience". This is a separate function of button B. A handout leaves the public-sector employee on the roll — he will still be brought in on Friday morning (§13d). Today he has three bad moves: vote under coercion, refuse and quarrel with the boss, or quietly click the wrong thing under the boss's eye. Exit gives a fourth: "I took the payment, by law I do not vote" — a record in the treasury, and the turnout ceiling does not let him be written back in (048f §3.2). It is conscientious objection, only paid. This state is fixed under the term "the meek" (011b §11b.2b).
Hence the exact answer to "we pay": a one-off payment and exit are different goods. The first is money with the bringing-in preserved, the second is release from the bringing-in. The question "where are my 30,000" in the mouth of a public-sector employee means "let me go", and a handout cannot close it in principle. The boss will forbid pressing B — then coercion costs the person a sum he can see (§13d); exit is not secret (048k §3), but with 40–50 % having exited they cannot be dismissed, and pressure on the choice is criminal (048k §4b).
Not a victory but an undermining. The architect: "talk among the people has already started — and is there much talk among the people about smart voting, and did the power ever publicly take notice of it? This is not a victory over the power but one more way to undermine it; perhaps it will be decisive, perhaps something more is needed". To smart voting the power replied in 2021 with repression and in silence — removed the app, blocked the site; in the core electorate it was not discussed at all. Bonya was answered in words and within a day (§13f). Repression answers a threat inside the count; words answer a question from one's own.
What more is needed. The question creates demand; what makes the undermining decisive is the supply side, and of its three parts none is ready:
- A live precedent — the pilot (048l), preferably outside Russia: before it the sum is a word, after it a fact any blogger will cite.
- The text of the rule on one page — the charter (048k §4): once the talk has started, people need something to point at.
- A window — the moment when contenders compete for the core electorate. The question does not open the window; it ensures that in the window the first contender picks up a ready demand, because it is already popular and already costed. An autocrat pays willingly (029.4b) — the question sets what for.
The question is not a lever but a charge that lies there and grows dearer; it will go off from someone else's spark.
13k. Not a patch: an idea with no expiry date, which the opposition lacks¶
The architect, 01.10.2026: "as a symbol, as an idea for the opposition the protocol is not a patch, not a temporary measure for these elections, not a 1–2 % tweak; it is an honest, clear and timeless protocol that changes the whole system, and it can be promoted for years and even decades from different sides. It is clear and needed by the people from any side. It is the idea that unites, and the opposition has no such idea now".
The opposition's whole repertoire is tactics with a date: smart voting, "Noon", boycott, signatures. Each dies on polling day and is reinvented (§13i). Only "against" unites the opposition; on "for" it is split — who leads, which candidate, which programme (056d). The protocol differs in genre.
- A rule, not a person and not a party. It does not require agreement on who is in charge or what the policy is: "the right, not the methods" (048k §3). The subject of the quarrels is taken out of it.
- No expiry date. Explanations accumulate over years rather than burning out on 21 September.
- An entrance from any side. For the left — a dividend to the poor; for the liberal — a filter against populism; for the patriot — an honest number for the top (§13g); for the public-sector employee — an alibi (§13j); for business — the thermostat. One idea, five doors, and none requires changing camp first.
- An answer to "and what do you propose?" — the question on which the opposition has been losing to the core electorate for twenty years.
There is no name in the protocol — and that is a race, not an obstacle. The first version named as the weak point: an idea without a leader gives leaders no reason to carry it. The architect: "they will find where to write their name in; but not to hitch oneself to such a rocket is to make a present of it to a competitor; nobody is going to ask their will here". Correct: the idea travels in the core electorate without them (§13i), and each has only the choice "with it" or "past it". The name writes itself in — "I said it first", "my calculation of the sum", "my bill" — and the protocol does not change for it: the right, not the methods.
Where the name is written in — in the parameters. I had recorded as the weak point the idea splitting into versions. The architect: "they will write their name in during the debates on the protocol's details — what per cent, how to count, when and where; we have left a great deal for the experts. There will be differences there, and there is room to roam". The charter deliberately left open the per cent of the median, the calculation of the median by country, the late-payment discount, the time scale, the levels of election (048k §3) — that is the place for names: "the sum according to one", "the formula of another". The core — two buttons, payment for exit, the weight of those who remain — is not open to dispute, because it is the idea. Differences in parameters do not split it but advertise it: every dispute about the per cent is one more video about the protocol.
The surgeons did not leave. The architect: "after the invention of penicillin the surgeons did not leave — operations simply became more complex and more interesting". This answers the politician's main fear: the protocol does not abolish his profession. Before penicillin a surgeon spent half his effort on the patient not dying of infection; after it he began doing operations nobody had thought of. A politician today spends his effort on turnout, bribery, observation and the fight over the count; after the protocol the dirty part is gone, and the work itself remains and grows more complex: persuading those who remain, arguing over parameters, running projects several cycles long (013f §4b, 040k §3).
Whose fear is removed and whose is not. The architect: "this removes the fear of the clever politicians, and the stupid ones ought to be afraid. Today they will always outshout the clever — Zhirinovsky and Trump are a very clear example". The mechanism is that of 023c §5: the audience sets the politician. The loud one beats the clever one not by strength but by shouting to an audience that has neither time nor reason to listen. The protocol does not silence the loud one — it leads his listeners away with the money. Those who remain have paid for the right to decide by forgoing the sum and listen to the end (013f §4b): argument works on them, shouting does not. The one who should be afraid is the one whose only tool is volume.
Weak point. The border between core and parameters is nowhere written on one page: until the charter exists (048k §4), the core itself may be declared a "parameter" — for instance, paying for turnout rather than for exit.
13l. Peaceful Russia: what the party plans on the ground and whether it needs the protocol¶
The architect's question of 01.10.2026: "find on the site what they plan to do on the ground and compare their plans with the protocol; perhaps they need it badly?" Read: the manifesto, the charter (048m), the Central Political Council's statement of 21.09.2026 on the election results, the "Noon" announcement of 12.08, the chairman's article of 08.08, the bureau's statements.
What they plan. Before the election — "Noon for a peaceful Russia": come on 20 September at 12:00, vote against United Russia, photograph the queues. After the election — the statement of 21.09: the Duma is appointed, legal opposition work in Russia no longer exists; the regime rests on its ability "to pay for loyalty and punish dissent", and that resource is running out; people can unite situationally but cannot hold positions for long; the electoral system that could secure a democratic transition is destroyed. The plan — by the window of opportunity to restore the capacity for organised action through a clandestine network in Russia and in emigration with its centre abroad; membership is anonymous. The rest of the work is addressed outside Russia: a letter to the European Commission on deserters, an article for European politicians, statements on political prisoners; at the congress the lines of work named were education on the internet, support for supporters inside the country and contacts with the West.
| Peaceful Russia | The protocol | |
|---|---|---|
| Diagnosis of the regime | pays for loyalty, the resource is running out | the same: a monopsony on loyalty (§13j) |
| What is awaited | the window of opportunity | the same window (§13j) |
| What will be in the window | an organised force; which electoral system is not said | a ready rule the first contender picks up as a demand (048m) |
| Action now | a clandestine network: membership, a centre, risk | a question without membership, centre or risk (§13i) |
| Audience | the anti-war minority and Europe | the core electorate, an entrance from any side (§13k) |
| Tactic at the election | "Noon": the queue is visible, there is no number | a sign on the ballot: a number in the precinct protocol (§13i) |
| Lifespan | until polling day | none |
Do they need it — yes, in three places, and these are their own blanks.
- "What comes after". They themselves write that the transition is not guaranteed and the electoral system is destroyed; the manifesto has one line on it, about honest and transparent rules. The protocol is the content of that line.
- A message for the majority. Their programme requires the listener first to agree about the war; the protocol does not. Without the core electorate the window will open for someone else.
- A safe task for the network. A network without a task is a list of names. "Ask in your own bubble where my 30,000 are" is an assignment that is lawful and safe, but not because the participant is unseen. The architect's correction: "it does expose the participant, but he may just as well be for the power, and if asked he will confirm exactly that". The protection lies not in secrecy but in indistinguishability: the same question is asked by the power's supporter, its opponent and the indifferent alike, so no list of friends and foes can be drawn from it.
The protocol does not care who raises it. The first version said here: the party is dangerous as a face, the label will stick to the question, "unfit as a carrier". The architect: "the protocol does not care who raises it. The Pythagorean theorem does not care whether they are bad barbarians or good Romans: it gives knowledge and possibilities and works for everyone. Nothing unwanted will stick there except popularity and interest. There are risks, but the overall result will always be positive — even if Putin personally starts pushing the protocol, it will be good". Three reasons why this is right.
- A label sticks to a person and a party, not to money. Smart voting was killed by a label because it was an instrument for the opposition's candidates — the label hit the content (§10). The content of the question "where are my 30,000" is the voter's money; nobody refuses his own money because of who mentioned it first. Bonya was attacked as a person, the facts stayed facts, and the second wave was larger than the first (§13g).
- A rule works for everyone. If a party in exile raises it, the party gains popularity and the public-sector employee gains exit (§13j). If the power raises it, it pays by the rule, on the median, with a counter that cannot be drawn, and lets go those it brings in today: this is the trap for the autocrat of 039. Each has his own intention; the mechanism does its own work.
- The sign does not depend on the carrier. Only the speed does: a labelled first voice may hold back the loyalist channels for a time; it cannot stop them.
The only thing left of "will not fit" is on their side, not ours: their identity is values, and payment for exit may look to them like bribery (029).
13m. Gallyamov, 02.10.2026: four places where his analysis meets the protocol¶
Source: Abbas Gallyamov's interview with the Golovanov channel, 2 October 2026 (automatic transcript). The conversation does not mention the protocol; what follows is the assistant's reading at the architect's request: what in it serves as a contrast, what as an analogy, and separately on the United States.
1. "Silence is consent." Gallyamov: most Russians have already understood what is happening, "but try to come out with it — you become an enemy at once… Russians keep quiet, and Putin takes their silence for consent". That is an exact description of the position in which the protocol finds its person: understanding, dependent and silent (§3c; 011b §11b.2b). The contrast is one. Every present way of not keeping quiet requires "coming out": saying, signing, going into the street. Button B is the only known act in which silence stops reading as consent while nobody comes out: the person says nothing, signs nothing, and the payment taken enters the counter. The authorities are left to read not silence but a number. This is not a victory over the authorities, nor an argument that Russia will adopt the rule: that only a question is proposed now is in §13i.
2. Lying as a form of courage, and subordinates who stop reporting. Gallyamov describes a system in which a general finds it safer not to report a failure and a superior not to listen, and the lie turns from a political problem into a military one: the neighbouring unit goes where the report says "ours". That is an analogy for 048e and 048f: the only defence against such a system is a number that depends neither on the reporter's courage nor on the superior's wish to listen. The counter by precinct is published every day and "may not be suspended on any grounds" (048m, Article 9). The reconciliation of two records (048f) is built precisely against a report in which "taken" is written before it is taken. The caveat is the same as in §13e: a rule adopted where there are no institutions is protected no better than any other law; its strength is that it exists before the moment the authorities wish to disregard it, and in the price of disregard, which everyone sees.
3. The people want a boring government. Gallyamov: the ideal for the overwhelming majority now is "something like Brezhnev", dull and predictable, "you open the paper in the morning and know for certain there will be no news in it"; the authorities' sharp phrases are a sign that "it will not be boring". That is an analogy for the design of the protocol itself, not for its Russian part. The landing page and 001b say the same in other words: the right not to be interested in politics is recognised as lawful. The protocol is a boring construction: one question, two buttons, a payment by formula, and none of it requires a movement. Whoever wants Brezhnev is not an opponent of the rule but its quiet voter: he chooses B and gets exactly what he wants — to be left alone, lawfully and for a payment.
4. Sanctions as a non-renewable resource, and converting a sanction into a number of drones. Gallyamov proposes "to convert every sanction into a number of drones" and further into a number of Ukrainian lives, so that a deal lifting sanctions becomes visible in units nobody can fail to understand. That is the method of §5 of this chapter ("the exchange rate": how many dividends in one line of luxury) and of 039 on Hungary: turning the price of a political decision into a number visible to everyone. The protocol does this permanently, not on occasion: the price of bad government is written into the size of the payment and the number who take it.
On the United States — separately. The most useful place in the conversation for the American part of the repository is not about Russia. Gallyamov explains why Congress cannot "lift sanctions today and impose them tomorrow": "congressmen have to get elected, they cannot afford to look like weathervanes; in each one's district competitors from his own party and from the other watch him, they file away everything you do, and you simply lose the next election". That is the institution on which 018 §34 and the whole of Part IV rest: not the virtue of the elected but the competitor in the district who remembers. Two conclusions, and the second matters more.
The first is a confirmation: where such competition exists, zigzags are impossible, and a rule adopted by referendum stands on the same support as the sanctions of Congress: it can be repealed, but everyone who voted for repeal will have to explain it to those who were counting their money (048g).
The second is a boundary. Gallyamov's mechanism works where the next election can be lost. In a district where the seat is safe and turnout at a midterm is 20 %, the competitor from the other party does not watch the congressman, because he has nothing to threaten him with; only the primary watches, that is, the core. The repository asserts exactly this (021; 059e): competition died not in Congress but in the districts where the composition of those who turn up is set by a machine. The protocol is needed not where Gallyamov's institutions work, but where they exist on paper and do not work: it gives the congressman back the very competitor who "files away everything" — through voters who for the first time have something to lose from his vote, because their money depends on how he votes on the budget.
The architect's question, 03.10.2026: "will the generals, who can no longer lie themselves, shoot at a people that asks for AB-EXIT as money, if Putin orders them to?" The argument cannot rest on "nobody shoots at people asking for money": Novocherkassk in 1962 was precisely about money — prices were raised, piece rates cut, the workers came out, and they were shot at. The strength of the question lies elsewhere, and Gallyamov described it himself: the screws are tightened on those who "come out", and here nobody needs to come out. There is no crowd: the question is asked in kitchens and in comments, and the anonymous count is collected by an application, not on a square (§7). There is no addressee: it is asked not by the opposition but by the regime's own base — the teacher on the commission, the state employee, the pensioner (§3c); one does not shoot at one's own base, not out of humanity but because one stands on it, and pressure on it is already applied. There is no action against which an order can be given: the person signed nothing and went nowhere, and to ban a question about money is to admit aloud that the constitution's people as the source of power is a fairy tale (§7). And the generals are beside the point, for a different reason than the assistant first wrote. The architect's correction: "it is their own problems they will not report, whereas here the people want money — more exactly, to go home quietly with the money". Gallyamov's "they will not report" concerns their own failures: a general hides a town not taken. The people's wish to go home with the money is nobody's failure, and nobody needs to report it: it is not a complaint that must reach the top but an act performed below. The counter does not replace a report — it only records what people have already done. The boundary of the argument is the same as in §7: all this concerns the question, not a legal collection of signatures with passport data; there the risk is high and personal, and the regime acts as usual. The architect's refinement the same day: "and the soldiers of those generals? The people come out to a rally for AB-EXIT — will they be dispersed, beaten, shot at?" They will be dispersed. The content of the slogan does not stop the police or the army: Belarus in 2020 — hundreds of thousands with the most universal demand possible, "count honestly", and every riot policeman knew how his mother had voted; they beat people. A man in uniform does not weigh what people came out for; he weighs who will win: while the vertical stands he beats, when it shakes he steps aside, as in August 1991. That is why there is no rally in the protocol: not out of caution, but because the authorities have an answer to a square and none to a question from a kitchen or to an anonymous count. Where the soldier and the policeman matter for the protocol in earnest is on the ballot: they and their families are the base that is brought in; under the rule the riot policeman takes the payment like the teacher, and in a secret vote on his own money he votes "yes" (048n, the fork for the machine). The protocol does not ask the man in uniform not to beat — it makes him a recipient. What follows from the question is not "they will not shoot" but "there is nobody to shoot at, and they will vote".
What must not be taken from this conversation. The assessments of the front, of oil refining and of the public mood are the words of one analyst, which the assistant has not checked; they do not enter the repository. On Israel and the aircraft — nothing, except one observation: a culture in which passengers run to the cockpit rather than sit belted in is not a condition of the protocol; button B is written for those who stay seated.
13n. The main question: the authorities will understand and send the men in uniform — can they¶
The architect's question, 03.10.2026: "The authorities will know for certain that the payment is dangerous and important not for its sum but for the exit from voting, and will be categorically against it, but will not be able to answer openly — and will send the police and the army to suppress and to coerce. Can they?" The repository has the answer, but scattered over five sections; here it is gathered.
What the authorities can do. Disperse any gathering — Belarus 2020, §13m. Imprison the organisers of a legal collection of signatures — which is why §7 separates it from the anonymous count and calls the risk high and personal. Force state employees to vote — this is done already, with tracking by name and a family multiplier (§3c). The protocol cancels none of this and promises to cancel none of it.
What they cannot do — three things, and the protocol is built on them.
- Suppress what has no organisation and no addressee. The question "the parties are paid for my vote, and I am not" needs no carrier and spreads by itself: pro-regime bloggers, state employees and pensioners raise it, and the protocol does not care who raises it (§13g, §13l). Riot police are not sent against the audience of Z-bloggers or against one's own teacher on the commission. Belarus was about a crowd; here there is no crowd.
- Answer. The authorities will know that the danger is not the sum but the exit — and for that very reason cannot say so aloud: "we need your brought-in votes for free" is a sentence that cannot be spoken. "There is no money" — impossible: deputies are paid, and it shows (§13h). To ban a question about money is to admit that the people as the source of power in the constitution is a fairy tale (§7). Silence remains, and silence in reply to a question about money works for the question (§10).
- Make coercion free. Today the brought-in voter votes and loses nothing — he simply came. After the question the same person knows he was made to give away for nothing what the parties are paid for. Coercion does not disappear, but it acquires a price, paid by the authorities: the compressed spring (011b §11b.2c) and the bill on every branch (§3b). Hungary in 2026 showed that the spring releases at the first opportunity — plus nine points of turnout (039).
The architect's refinement, the same day: "a gathering for the greatness of Belarus — and, following from it, the payment to the people under the protocol. On what basis, logical rather than legal, is it to be suppressed?" There is no logical basis, and that is the answer. A legal one will always be found — "unauthorised" — but the authorities have nothing to tell themselves or the people about why they disperse people with the state flag who ask for a payment under law: "patriots asking for money are enemies" is a sentence of the same kind as "we need your votes for free". Belarus in 2020 could call the crowd foreign agents; a crowd with the slogan "pay the people" cannot be called that, and what remains is "provocation", a word the people have heard before. The right formulation is therefore not "they cannot" but "on every branch the authorities pay" (§3b): to disperse is to burn one's own base — those who came for greatness and got the truncheon join the burned, and that is the compressed spring; to tolerate is to recognise the question as lawful, after which it travels by itself; to promise money is the authorities' bow, which under §13j is not the protocol's loss but its work. The assistant first qualified: "the authorities disperse without a basis and invent one afterwards". The architect's correction: "the authorities always disperse with a logical justification, otherwise they would ban breathing, and that is illogical; but with AB-EXIT it will be harder". That is more exact. The authorities always have a basis — the difference is which one lies to hand. Against a crowd with foreign flags it is ready: foreign agents, extremism, enemies. Against people with the state flag asking for a payment under law there is none ready — it has to be invented, and each invention ("provocation", "bought by the West to ask their own state for money") is worse than the last. Not impossible — harder, and that difficulty is the price: the argument is not that dispersal is impossible but that here the justification for it costs them more than the question they wanted to close. The protocol proposes no gathering; but if people come out in such a frame, the zugzwang is the one the front-runner faces at a debate (001c): every move is worse than inaction, and inaction is a move too.
The boundary. The protocol does not claim that the question will bring the authorities down or that they cannot hold on by force. It claims three things: there is nothing to suppress, nothing to answer with, and coercion grows dearer with every cycle. While the vertical stands, it stands. But the idea does not wait — it sharpens. The architect's correction, 03.10.2026: "the opposition sharpens the idea and keeps it up at every election: had you adopted AB-EXIT, tomorrow you could take 800 dollars — and what did you get at the election? Think about it". Every election is an occasion to present the same bill, and the bill grows: over a cycle a person sees how many times he gave his vote away for nothing. That is what the share calculator (§5) is written for: it counts not a payment in general but the one this person did not receive at this election. The vertical need not shake for the bill to grow; when it shakes, the question lies ready in everyone's hands, including those of the men in uniform, who are themselves its recipients.
13o. The arguments in her head: a woman of sixty-five, alone, who came out for her share¶
The architect's question, 03.10.2026: what arguments to the authorities are in the head of a woman of sixty-five without a husband who came out for the protocol. And his correction to the first version: the comparison "the parties are paid for the vote" is weak — there is little money in it (152 roubles a year); the main line is "the country is rich, all is well, and from that good a small share to the people: the authorities have enough, the deputies have enough, and the aunt is given nothing". Everything below is written in that frame: she believes the television, and every argument of hers grows out of what the authorities themselves have said. A pair to §13h: there the arsenal is for the opposition, here it is for her.
Who she is. Sixty-five, alone, a pension of twenty-odd thousand. For forty years she slaved at the plant, and for forty years she went to vote as she was told. She watches television and believes it: the country is rich, rising from its knees, the enemies envy it. Her flag is her own, not issued.
Who she is by the data: Putin's core and Orbán's core. The architect's question the same day: what is the core electorate of Putin and of Orbán, give a portrait. The portrait coincides with her almost word for word. Levada: the typical voter of the ruling party is a woman of 55 or older, with secondary vocational education, middle or below-middle income, in a small town or a village; trust in Putin is higher among the old, women, people with secondary education and modest income, in villages and small towns; by dependence on the state — public-sector employees, workers of state enterprises, pensioners, "the social and cultural periphery" (Gudkov). The most loyal group in every poll of 2025–2026 is those who trust television: 83–95 % approval; among the 55+, 76 % in June 2026 while overall approval fell from 82 % to 74 % in half a year. The affluent and Moscow approve no less (80–88 %), but they are not brought in by a call from the clinic; the first group is. Orbán: pensioners and older villagers; in settlements of up to 500 inhabitants Fidesz held 50.3 % against Tisza's 41.9 % even on the day of defeat; before the election the government itself named its core in money — grants of 3 million forints to pensioners in villages of under 5,000; among voters under 30 Fidesz had 15 %, Tisza over 60 %. And that core moved first: Fidesz lost more than 200 thousand votes in rural communities, and the villages, its former bastions, swung to Tisza most of all. Hence the conclusion for the whole section: a woman of sixty-five from the plant, with a television and her own flag, is not an invented character but the statistical centre of the core of both regimes. The protocol addresses not the opponents of the authorities but their foundation, and Hungary showed that the foundation moves when it is presented a bill in money: Orbán paid it grants, Magyar promised to bring back the frozen billions, and it chose the second (039; §13b). Sources: Levada Centre ratings of November 2025, April and June 2026; L. Gudkov's interview on the erosion of the core; Átlátszó and LSE EUROPP on the results of 12 April 2026 by settlement type; the report on rural grants before the election. The Hungarian figures are from secondary analyses; the assistant has not checked the precinct protocols.
Her starting thought is not a grievance but a conclusion from the news. "You say so yourselves: the country is rich, all is well, the oil is ours, the budget is a record. I believe it. So from that good you have done for the people, the people can be handed a small share. The authorities have enough, the deputies have enough — and the aunt is given nothing. Forty thousand once in five years. I am not asking for anyone else's. I am asking for mine out of our common good."
| What the authorities will say | What she answers to herself |
|---|---|
| "There is no money" | How is there none? On television there is. Or does the television lie? Then say it lies. |
| "This is populism, a handout" | A handout is when everyone gets it. This is by law, once in five years, and only to those who took it. A hundred and fifty thousand a month for deputies' lunches — is that not a handout? |
| "This is vote-buying" | Vote-buying is a bag of buckwheat for a tick; I have seen it. Here it is the same for everyone and no matter whom you vote for. It is a share, not a bribe. |
| "The West invented it to shake the country" | The West offered me nothing; it is my enemy, you said so yourselves. I am asking my own about ours: where is my share of our good? |
| "The vote is sacred, nobody pays for it" | Sacred. So I am not selling it. I ask for a small share of the common good, and the vote stays with me: if I want, I go. |
| "You are against the state" | Forty years slaving at the plant and forty years for. For now too — here is the flag. I ask that the state, which is doing well, do well by me too. Is that against? |
| "Extremism, unauthorised" | I am sixty-five, in a headscarf with a flag. If I am an extremist, who then are your own? |
| "The elections are honest, why change" | I change nothing in the elections. The elections are yours, the count is yours. I ask for a share — and if I am not needed, I will sit at home. If I am needed, the share is earned. |
| "Go home, granny" | I will. With the share. My son at the plant on forty thousand will take it too. And my granddaughter the teacher, whom they put on the commission. |
When the authorities turn to the people. That is their main move: not "the idea is bad" but "the people are bad" — at the carrier, because there is nothing to strike the idea with (§13g).
| What the authorities will say | What she answers to herself |
|---|---|
| "Foreign agents and runaways propose this" | I do not need the people, I need the share. Lock them up — and give the share yourselves. It will be no worse. |
| "It is bait" | Bait has a hook. Where is the hook? Money from our budget under our law. Show me the hook. |
| "They want you not to vote so that the government falls" | I do not vote anyway — I come because I am told to. If the government falls without me, it stands on me. Then the share is earned. |
| "Money today, and tomorrow they take your vote away for good" | The rule says: at the next election I choose again, and repeal is only by the same vote of everyone. Without the rule nobody asks me at all. |
| "They want a minority to decide for you" | Those who brought me decide for me now. Let those who came by themselves decide. If I want to decide, I do not take it; nobody takes that right away. |
| "They are destroying the country" | The country is destroyed by those who are doing well and give the people no share. I am for the country. |
| "Normal people do not ask for such things" | Bonya asked — three million watched (§13f). She is no foreign agent. Or is she, by now? |
The last argument, to which there is no reply. She demands neither a change of power nor an honest count. She demands a share of the good the authorities themselves speak of every evening. Any reply is either "there is no good", and then the television lied; or "there is, but not for you", and then for whom; or "we need your vote for free", and that will not be said. The authorities can take the idea away from the "bad people" in one way only — by proposing it themselves, and having proposed it, they have bowed (§13j). Silence remains (§10) — and silence, in the architect's correction, "is good while people wait and do not know where to ask": it shelters until the opposition tells the people aloud what it means. That is the division of labour of §13h: the question spreads by itself through the loyalist milieu, and the opposition names the answer the authorities cannot speak and repeats the bill at every election (§13n).
And the policeman. The architect: "now imagine the policeman telling her all this — and how he starts to beat her with such inner logic". That is the test of the whole construction. To disperse her he must first say one of these "no"s to her — and she has just taken each of them apart, and he heard it. His mother is her age and from the same plant. He will not refuse the order; §13m is honest about that. But the architect's correction: "orders to kill are given rarely; what is given is an order to suppress — that is, to prevent lawlessness". And here the order runs into its own word: there is no lawlessness, she is asking for a law. To carry it out the policeman must first name the lawlessness he is stopping — and before him stands a woman with the state flag asking for a share of the common good under law. And the grandmothers say "for the Motherland", and the policemen say "for the Motherland", and everyone can hear that both are for the Motherland. At that moment the authorities run out of what the order "suppress" rests on — the monopoly on the word "Motherland". He will raise his hand; but a scene in which both sides shout the same thing will have spread before he lowers it, and the second such order will cost more than the first. And on the ballot he is a recipient, like her. The architect: "and here the free-rider law is already on the grandmother's side, and the policeman is a free rider for the protocol". That inverts the free-rider objection of 048n: there it was said that a person will not go to the referendum because the money will come without him. Here the same law works for her: the policeman need do nothing for his share — she is standing for it. The cheapest action of a free rider is inaction, and inaction in his case is a refusal to suppress. He rides on her ticket. And not he alone — the architect: "his whole family and relatives, six of them; his parents and his wife's parents ride there too: at least something for them from the state". That is the family multiplier of §3c — the one the authorities coerce with, when a mother on the state payroll talks her whole family round — only in reverse. His wife, his parents, his wife's parents: six people at forty thousand each, and for the old people on both sides it is the first thing the state has given them personally in their lives. The order "suppress" is given to him alone, and he counts for six. And one more thing, from the architect: "it is not a special squad on a big salary standing there, but very ordinary policemen". It is not an elite unit on a Moscow wage that is sent to a grandmother with a flag — it is the district officer and the patrol, for whom forty thousand a person is nearly a month's pay. For him the share is not an abstraction but his own wage multiplied by six; and the authorities will not find a special squad for every such square.
"A small share" is not a figure of speech: forty thousand for 55–66 million takers is 440–530 billion roubles a year, about one per cent of the budget, four kopecks of every oil-and-gas rouble (§5, the third scale).
13p. How the authorities will argue that the protocol is evil, and what to answer¶
The architect's question, 03.10.2026: "how will the authorities convince people that the protocol is evil, and what can we object logically". Below are not the arguments the authorities will throw at once, but those a clever spin doctor would write for them for television — by strength, with a logical reply and a third column: what remains, honestly. Three arguments stand on real gaps of the repository, and they must be met with admission, not denial.
| The authorities will say | The logical reply | What remains, honestly |
|---|---|---|
| "If there is money, we will raise pensions for everyone and every month, not a handout once in five years." The strongest argument: loyalist and generous. | Pensions could have been raised all these years — they were not. A pension rise is the authorities' discretion, given and taken away; the payment is a right by referendum that the authorities can neither set nor cancel. One does not hinder the other: raise the pensions and give the share. And if "we will raise" — raise, we will wait with the figures in hand. | The authorities can indeed answer with a pension rise — that is their bow (§13j), and for the protocol it is not a loss. |
| "This is vote-trading. Our grandfathers fought for the right to vote, and you are selling it." | Nobody sells anything: the vote remains, it can be used at any election, and it cannot be taken away. The grandfathers fought for the right to choose — including the right not to go. Trading is when a candidate buys a vote; a candidate is forbidden to pay by the charter (048m, Art. 1). | — |
| "The poor will take the money, the rich will vote — the rich will rule the country." An argument with a core of truth. | The architect's correction: "are you afraid of the rich or of the people? The people are not afraid of the rich — so why are the authorities?" The aunt is not afraid that the doctor, the engineer and the shop owner will vote: they are her children. The authorities are afraid of them — because they have a stake and a head of their own, and a call from the clinic will not raise them. "The rich will rule" in the mouths of those who rule is an admission: they fear not the rich but those who cannot be brought in. And under the rule the poor keep the vote and return at any election; the rich man does not take forty thousand not because he is better but because it is too little for him. | Selection runs partly by income — the repository has neither proved nor refuted this (040m §4). It is a real weak point, and the authorities will find it. |
| "Nowhere in the world does this exist. Do you want to be guinea pigs?" | The architect: "everywhere in the world the people are the source of power; so we, as the power, want a part of the country's profit". That is what your constitution says. We are not asking for an experiment — we are asking for an owner's dividend. And we have already been made guinea pigs: electronic voting was introduced without a pilot across the whole country, and it is used for coercion (§13d). A pilot — a courtyard, a village, a town (048l) — is needed not to test the right but to find the sum. | Introduced nowhere; no pilot has been run. That is a fact, and we do not hide it. |
| "Today they pay you for not turning up; tomorrow they cancel elections — and nobody notices." | The payment is tied to the calendar of the cycle, not to the fact of an election (048m, Art. 6): cancel the election — pay all the same. The authorities can cancel elections today, for free. | — |
| "The treasury will not bear it. They start with forty thousand and will demand a hundred. The budget bursts, and you will be to blame." | Forty thousand for 55–66 million is about one per cent of the budget a year, four kopecks of every oil rouble (§5). Only a referendum can raise the percentage, and before it those who pay taxes see the bill (018 §34). A pension is the same kind of obligation, and the treasury bears it. | What it really costs nobody knows: the percentage is not chosen, the share of takers is an estimate. |
| "Turnout will fall — the enemies will say the government is illegitimate." That is the authorities' real fear. | The architect: "if the people need this government, they will not take the handout but will proudly go and vote. Or do you doubt the people?" Any reply is bad: "we do not doubt" — then what is there to fear; "we doubt" — and there, it is said aloud. And legitimacy is not the number who came but a number that cannot be drawn: the counter shows how many took the payment and how many voted, and everyone accepts it. Eighty per cent turnout, half of it brought in, convinces nobody even now. | — |
| "They are buying your silence. They will pay — and do what they like without you." | The architect, in the aunt's voice: "What, did you just call me a bought beggar? I will not take it — I am for the referendum. My neighbour and the old man need it, and I am a hundred per cent for the government, I will go and vote." The accusation misses: the rule is voted for not only by those who will take the money — that is the second group "for", which Grok did not see (048n). It is not she who is bought; she asks for her neighbour and the old man and stays to vote. And one who took the payment is not silent either: he keeps the courts, protest and the referendum on the rule itself (Arts. 5, 16), and "what they like" shows on the counter — when eighty per cent take the money, that is the scream. | — |
| "A foreign project; the author is a nobody from Chișinău with AI chats." | The protocol does not care who raises it (§13l). If you dislike the author, take the idea and propose it yourselves — it becomes yours. A rule is checked by its text, and the text is open in full. | There is no external review; the text was written with AI. We say so ourselves, before they do (1e). |
| "It is humiliating: the state pays citizens not to take part. An insult to democracy." | The state already pays for taking part — the parties, the contractors, the public-sector workers with a day off for turning up. Humiliation is being taken by bus. Here a person chooses himself and is paid for the honesty of the choice. | — |
| "It splits the people: those who vote and those who were bought." | The split exists now: those who were brought in and those who stayed home and do not count. The rule counts everyone for the first time — those who took, those who came, those who chose nothing. Nobody is "bought": the choice is made anew at every election, and the neighbours do not know who chose what (Art. 9). | — |
Where victory in the argument cannot be promised. Three arguments stand on real gaps: "it exists nowhere" is true, and the only answer is a pilot; "selection by income" is untested, and the pilot measures exactly that; "what it costs" — the percentage is not chosen, and the answer is a table, not a promise. Authorities who say these three things calmly, without shouting, are stronger than us on air. The protocol's advantage is not that it has an answer to everything, but that the authorities have no answer to one question — "if my vote is not needed, pay and let me go; if it is needed, it is worth money" (§13o) — and the woman of §13o turns every argument in this table into it in one line.
13q. One electorate for two: an exact portrait and a fork for the authorities¶
The architect, 03.10.2026: "Orbán and Putin have one and the same core electorate. Describe it exactly — and we will think how to show it the protocol so that the authorities fall into a fork". The portrait below is from the same sources as §13o; where it is inference rather than data, this is said.
- Who. A woman (about two-thirds of the core), 55–75, centred on 60–70. A village or a town of under 50 thousand; in Hungary a settlement of under 5 thousand. Secondary vocational education: a technical school, a trade school, a nursing college. Data.
- What she lives on. A pension or a wage from the budget: school, clinic, post office, village council; in Hungary, public works from the mayor. Income below the median, and the money comes from the state directly — she knows it every month when it arrives. Forty thousand once in five years is for her one and a half to two pensions. Data on composition; an estimate of size.
- With whom. Alone or with a pensioner husband; the children in the city, the grandchildren in summer; a vegetable plot. She helps the children, not the other way round. The son — a plant, security, the police; the daughter or granddaughter — a teacher, a nurse: the budget too. Inference from the composition.
- Where she learns about the world. Television is the main and almost only source; this group is the most loyal in every poll of 2025–2026 (83–95 % approval). The second channel — neighbours, the church (stronger in Hungary), the village chat. She has no independent sources and does not look for them. Data.
- How she relates to the state. The state is provider and boss at once: she obeys it and expects from it. She votes as told: in Russia a call from the clinic or a social worker, in Hungary the mayor on whom the public works depend. The key fact: she has already been paid before elections — ten thousand to pensioners before the Duma election of 2021, a thirteenth-month pension and a tax refund before the Hungarian election of 2022, grants of 3 million forints to rural pensioners before 2026. She knows that before an election they give; the authorities themselves taught her that. Data.
- What she values. Order, "that it not get worse", respect for superiors; the country — sincerely. She fears chaos, remembers the nineties; she regards the opposition as strangers — urban, rich, young. She wants not so much money as to be noticed: "we worked for forty years". Inference; supported by §3c.
- What she says and what she thinks. Aloud — "for", always. To herself — "they stole everything", "everything to the deputies, nothing to us", "nobody asks us". This is not oppositionism but a grievance with nowhere to go — the meek (011b §11b.2b). Inference.
- What she fears personally. That the supplement will be taken away, that prices will rise, that her son will be taken, that she will be left alone. And that she will be called an enemy — for her that is worse than poverty. Inference.
- What Hungary has already proved. The core is not a monolith and not an ideology: in April 2026 it was precisely the villages that swung the most, when they were shown the bill — the frozen billions against Orbán's grants (039). It is loyal until it is presented with a number. Data.
A refinement for the United States and Germany. There the portrait coincides in age and in television, but diverges in the main thing — the relation to the state. The US, 2024: turnout among the 65+ was 74.7 %, the highest and the only one to rise against 2020; among the 18–24, under half; the decider is a retired homeowner who comes by himself: nobody drives him and nobody calls him. Germany, 2025: the highest turnout is among the 50–69, about 85–87 %; voters of 60+ are 42 % of those entitled to vote; CDU/CSU and SPD get their best result among the 70+. The burned there are different people: the AfD's core is men of 25–44, the East, workers; they have not left the election but come to vote against. In Russia and Hungary the core is dependent and brought in, and it can be released for a share; in the US and Germany the core is reliable and voluntary, nobody holds it, and the lever "you are made to do it for free" does not apply to it. There the protocol's question is different: will the reliable but indifferent voter take the payment for what he no longer cares about — and will those who have something to lose remain to decide (055c §6.7; 018, 045, 059e).
Sources of the refinement: US Census Bureau, Current Population Survey, November 2024 (P20-590); the representative electoral statistics of the 2025 Bundestag election (Bundeswahlleiterin, Heft 4; Statistisches Landesamt Baden-Württemberg).
Where the fork is in the portrait — the assistant's proposal for discussion. The strongest lever is point 5: the authorities themselves paid her before elections. Hence a question she need borrow from nobody: "Before elections you gave — ten thousand, a thirteenth pension, grants. So the vote is worth money; you showed it yourselves. Then give it by law and always, not when you need it." Four moves for the authorities, all of them costly: to say "the vote is worth no money" is to admit those payments were bribes; to say "it is worth money" is to pay by law, that is, to bow (§13j); to pay once more as a one-off is to confirm that they pay, and to let her compare the one-off "when they need it" with the lawful "when it is due to her"; to stay silent is to have the opposition read the silence aloud (§13o). How to show the protocol to this person further is for the architect to decide; here there is only the portrait and the lever.
13r. The barrel as the only answer: a short with thirteen thousand likes¶
The architect sent (07.10.2026) the short "Our Realities, 30.09.2026" (Mikhail Khan's channel, 73 seconds; in six days about 308 thousand views and 13.4 thousand likes by the page's figures) with one line: "the protocol should have been done long ago". The short is a staged sketch: a citizen interrogates an official — why the birth rate falls, where a young family is to live at such a mortgage rate, why the young leave, "taxes up for the people, new houses for yourselves"; the official answers "I only executed, I was told and I did, I am a pawn"; the ending is a gun: "I believe you — but my barrel does not".
What is here for the repository is three things, and none of them is about the short.
The list of grievances is the same as the aunt's (§13o) and the apparatus's (039 §88.9). The mortgage, the birth rate, the departure of the young, "taxes for the people, houses for themselves", "I only executed" — these are exactly the arguments the repository gathered over a week from other sources. The short does not invent them, it repeats them, and thirteen thousand likes are not a rating of the script but a gauge of pressure: that many people recognised their own list. It is the compressed spring of 011b §11b.2c, filmed on a phone.
"I only executed" — the free rider on camera. The official in the sketch says exactly what the policeman of §13o says: there was an order, I carried it out, not my fault. The short shows where that logic leads when the citizen has no other answer than a gun: to the scene the repository precisely does not want (no-titania: the protocol is not an instrument of overthrow, 040n §5). The only answer today's system gives a person with this list is to keep silent or to shoot; both are free, and neither counts.
The weapon that does not fire. 029 §29.11 records: at the moment of arbitrariness the authorities have the weapon and the person does not and should not; the protocol gives the one weapon that does not fire and cannot be confiscated without admitting why — exit with a payment. The short illustrates that line by contrast: this is what imagination chooses when there is no lawful exit. A like under a scene with a gun and a "yes" to the rule in a poll are different numbers, and one cannot be carried into the other; but the short shows the direction of the pressure more exactly than any poll, and it coincides with what the protocol proposes to move from the barrel to the counter.
Boundary: this is a staging, not an event; the view figures are a snapshot of the page on the day of recording; and the short's author said nothing about the protocol — his list is used as data on demand, not as support.
14. Weak point of the chapter¶
The principal limitation lies in where the material comes from. Every election figure other than the budgetary and legal ones derives from three analyses published on a single day: Maxim Katz, Abbas Gallyamov and Meduza's discussion. The assistant did not verify precinct commission data; the estimates of the "real result" belong to those analyses and are drawn from honest precincts and exit polls, making them estimates rather than measurements.
Second: survey data from Russia do not belong here, a rule introduced in the course of the work (§13). Where it has been applied retrospectively some earlier claims were withdrawn — but no full revision of the chapter against that rule was carried out, and residues of survey material may remain.
Third, and material to every practical section: two things remain without an analogue — a payment tied to the act of voting, and an elite split caused by a dividend demand (§13). Everything else rests on a measurement or an observable case; these two do not.
Fourth: the chapter describes Russia, where the protocol cannot be adopted (056d.8), so its practical part — §4, §5, §6, §7 — is work for a transition window and for measurement rather than for polling day. It cannot be read as a programme of action yielding results within any foreseeable term.