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.
Reform by Subtraction: How the Protocol Repeals the Levies Nobody Can Repeal¶
Chapter: 04 — Electoral Dynamics (paired with §19) File: 04_019c · v1 · 19 September 2026 Source: the architect's and the assistant's analysis of three examples — compulsory driving schools, cash registers at small turnover, supervision of street food (Moldova, Thailand, United Kingdom); the fact-checking and the argument are recorded in 015b.6c. Here is the answer to the architect's question: "why did we discuss this? how can AB-EXIT solve it in Moldova or in any other country".
How to read this chapter (note of 02.10.2026). The text contains wording that is easy to misread: "a x2 vote", "x3", "the weight of a vote" are the arithmetic of a share, not a multiplier: every ballot counts as one. The exact-answers sheet 1d and the charter 048m are in force.
1. Why discuss this¶
The cost of bad governance (§19) is usually counted in large lines: corruption, debts, bankrupt cities. But most of that cost is spread across hundreds of small norms, each too small for newspapers to write about and too profitable to someone to be repealed. The compulsory driving school, the compulsory cash register in a small shop, the compulsory certificate, the compulsory intermediary, the compulsory inspection. This is where a person actually meets the state — not in the budget but in the queue for a paper — and this is where the effect of the protocol will be felt soonest.
Three examples showed the common construction. In Moldova, to take the driving test one must buy a course of theory and 30 lessons — up to 8–10 thousand lei, even if the person can already drive; in the United Kingdom there are no compulsory lessons at all, and road mortality is among the lowest in the world; in Thailand a school is voluntary, and without motorcyclists mortality is half Moldova's. In Moldova a cash register is compulsory, connected to online monitoring, and some models must be replaced at the owners' expense by 30 September 2026; in Thailand a business with a turnover of up to 1.8 million baht does not register for VAT at all. Food in Bangkok is checked by samples from the plate, not by the seller's papers (015b.6c).
2. The anatomy of a levy: three conditions¶
A levy-norm lives for decades when three conditions hold at once.
| Condition | Driving schools | Cash registers |
|---|---|---|
| The recipient of the income is concentrated and organised | School owners | Sellers and servicers of registers, inspectors |
| The payers are dispersed and pay rarely | Each pays once in a lifetime | Thousands of shopkeepers one by one |
| The majority that decides elections does not care | "The majority that votes has no car and no licence" | The buyer does not see the price of the register in the price of bread |
This is Olson's classic logic: concentrated benefit defeats dispersed costs. The consequence for a politician is simple arithmetic. The promise "repeal" brings no votes (the payers are few, young or busy and the worst at turning out; the rest are indifferent) and creates an organised, well-connected enemy. The promise is loss-making, so nobody makes it. The norm has not a single public defender — and not a single opponent.
This also shows that the bottleneck is not knowledge. Checking each of the three examples took minutes from open sources; any ministry could have done it twenty years ago. The norms live not because nobody knows, but because none of those who decide has any interest in knowing.
3. What the protocol changes: four links¶
The protocol does not and will not contain a norm "abolish driving schools". It breaks the third condition — and through it the rest.
Link 1. Composition. The indifferent take the sum. Among those who stay, the share of people who pay levies personally and regularly is higher: the entrepreneur, the self-employed, the car owner, the parent. And those who have already paid are not indifferent: a person who could drive and still sat through thirty lessons remembers not the sum but the humiliation of pointlessness.
Link 2. The price of a promise. The promise "repeal" turns from loss-making into the most profitable one in a campaign: it costs the budget nothing, is verifiable the day after the signature and requires no faith in the candidate's competence — the same class as "I will stop stealing" (019b). For a voter who counts his own money it translates into a sum: 8–10 thousand lei per child is four to five dividends (033c.9c).
Link 3. A lobby without its free resource. The recipients of the income do not go anywhere, but they are thousands against hundreds of thousands who stay; they can no longer bring the indifferent to the polling station — those took the money — and each of them, to vote, must himself refuse the sum (015c.7b).
Link 4. Feedback. If those who stay do not repeal the levies and life does not get cheaper, the exit share does not fall in the next cycle, and everyone sees it: the thermostat (015b.6b) measures not promises but whether people came back. Today there is no such feedback at all.
4. Reform by subtraction¶
Building a good institution is hard: it takes a design, money, people, years, and the result is not guaranteed. Removing a bad one is a single line, zero budget cost, an effect the next day. It does not even require competence — only the absence of a reason to keep the norm. "Abolishing them requires simply writing 'Repeal'. There is nothing to think over or adopt."
This explains where the manifold effect of the first cycles comes from when the base is low (015b.6b): not from brilliant managers but from hundreds of lines reading "repeal", each of which today is unprofitable for any politician. A country with decades of negative selection is a country with the largest stock of such lines.
5. The filter: how not to repeal what is needed¶
Not every norm that looks like a levy is one. A filter of two questions is cheap and applies to any norm.
- Comparison. Are there ten or twenty countries where the norm does not exist, and is the result worse there — crashes, poisonings, tax collection? If not worse — repeal.
- Result or process. Does the norm check a result (a driving test; a sample of food from the plate) or a process (buy a course; obtain a permit; show a paper)? A check of the result protects the person and costs the one checked almost nothing. A check of the process feeds an intermediary. Keep the first, repeal the second.
Hence an exact formula to which the objection "you want to abolish supervision" does not attach: free entry plus a check of the result — instead of permits and inspection of papers.
6. What it looks like in Moldova and in any other country¶
Moldova, the first cycle after adoption. Candidates compete with lists: "a hundred lines of 'repeal' in the first hundred days", each with the sum it returns to a family or a shop. Compiling such a list by the filter of §5 is weeks of work, not years. What a person gets: up to 8–10 thousand lei and two to three months for everyone obtaining a licence; the shopkeeper — the cost of the device, its servicing and its forced replacement; the buyer — a price with nobody else's rent in it; a young person with a stove and a pan — the right to start a business without a permit, under a check of the result. If tens of thousands of people obtain a licence each year, the first line alone is hundreds of millions of lei staying in families (an order of magnitude, not a calculation).
Any other country. The three conditions of §2 hold everywhere; only the height of the base differs. In rich countries the same levy looks like occupational licensing: in the United States roughly one worker in four needs a licence against one in twenty in the 1950s, and the explanation is not that occupations became more dangerous but the same construction — concentrated incumbents against dispersed newcomers and an indifferent majority. There the effect will be smaller and slower; in countries with a low base — manifold and fast.
6b. Two kinds of effect: fast and slow¶
This whole analysis began with the assistant's claim that the protocol would have no fast effect: "x10 is a condition, not a fact", "the cautious estimate is +2–5 % of GDP as a level effect", "the first cycle has to be financed before any effect". The architect: "we discussed all this when you were asserting that there would be no fast effect from AB-EXIT." The correction is accepted: the caution was right for one kind of effect and was wrongly extended to both.
| Construction | Subtraction | |
|---|---|---|
| What it is | New institutions, investment, quality of decisions | Repeal of levies and fictitious control |
| What it takes | A design, money, people, competence | One line and the absence of a reason to keep the norm |
| Cost to the budget | Yes | Zero |
| When the result shows | Years; +2–5 % of GDP as a level effect (019.59.5) | The day after the signature |
| Who sees it | An economist in the statistics | Everyone — in the receipt, in the queue, in the wallet |
| Sign of the result | Depends on the quality of the managers | Known in advance: the money stays with the person |
The repository's cautious estimates belong to the left-hand column and stand. The right-hand column is fast: its only delay is the first election after adoption, not years. Moreover, it starts working before the election: candidates who need cheap verifiable promises compile their "repeal" lists during the campaign, and the very publication of such lists changes the behaviour of inspectors.
For a campaign this distinction matters more than any other. A person need not be promised GDP growth in ten years; he can be given three lines that will vanish from his life in the first year, with a sum beside each. That is the answer to the question "what do I get besides the dividend".
6c. A test on housing: Moldova and Thailand¶
The architect's question: compare the rate of housing completion and its price relative to wages; "and Thailand is a country all the neighbours want to move to, while Moldova is a depressed region everyone leaves". The result of the check is mixed, and it is recorded in full.
| Chișinău / Moldova | Bangkok / Thailand | |
|---|---|---|
| Purchase price | About 1,720 euros per m² at the end of 2025; a 50 m² flat is roughly 131 average net salaries (13.2 thousand lei), that is, about 11 years | Average unit price 6.7 million baht (2024); a townhouse at about 4 million baht is roughly 250 average national wages (15.7 thousand baht) or about 160 Bangkok wages; the price-to-household-income ratio is about 28 years |
| Price dynamics | +60 % in two years (from ~1,070 euros per m² in 2023) | On average +4.3 % a year in 2012–2025 |
| Bottom tier of rents | A one-room flat from 320–350 euros, that is, about half the average net salary | A basic studio at 2–5 thousand baht, that is, 13–32 % of the average wage |
| Completions | The housing stock grew by 6.8 thousand dwellings in 2023 (about 2.8 per thousand inhabitants); in 2024 completions, by the statistics, fell 2.1-fold — to 2.6 thousand flats and houses, 200.9 thousand m² | Developers only: about 146 thousand units a year nationwide, of which about 96 thousand in the capital region (8–9 per thousand inhabitants); self-build is not included |
What was not confirmed. Buying new housing in Bangkok is not cheaper relative to income than in Chișinău: by the formal statistics it is dearer. Bangkok is a megacity of eleven million with foreign demand, and its new-build cannot be compared with Chișinău's in Thailand's favour.
What was confirmed, and it is stronger. First, the bottom tier: in Thailand the market produces housing for a person on an average or low wage (a room for a fifth of earnings); in Moldova that tier does not exist at all. Free entry gives not cheap luxury but the existence of the cheap. Second, the supply response. In a country people move to, prices rise four per cent a year and the market is if anything oversupplied; in a country people leave, with a shrinking population, prices rose sixty per cent in two years while completions simultaneously halved. Falling completions with a soaring price is the signature of supply that cannot answer demand; in a free market a 60 % price rise would set off a building boom, not a slump.
The architect's question: "and what, besides corruption and hidden permits, can affect supply like this?" An honest list of causes that give the same picture with no levy at all.
| Cause | How it works | Does it depend on the quality of governance |
|---|---|---|
| Construction lag | A building is completed two to three years after it starts; the completions of 2024 are the starts of 2021–2022, that is, the shock of a war next door and a policy rate above 20 % | No: an external shock; easy to test — if the lag is the cause, completions in 2026–2027 should rise sharply |
| The price of money | A developer builds on prepayments and credit; with dear money there are fewer starts | Partly |
| Shortage of builders | Workers left for where pay is triple; a country people leave loses not only buyers but those who build | Yes, but indirectly: emigration itself is a consequence of the low base |
| The price of materials | A jump after 2022, imports | No |
| Networks and land | The city lacks water, sewage and electricity capacity; connection is a separate queue | Yes: this is precisely the quality of city management |
| A few large developers | With few players it pays to hold supply back | Yes: players are few where entry is dear |
| Demand not from local wages | Diaspora money and buying "as a store of value" detach the price from wages | No; but this explains the price, not the fall in completions |
The conclusion from the table is twofold. The slump of 2024 specifically is in all likelihood explained in large part by the lag after the shock of 2022, and attributing it to permits would be dishonest. But the level — why in calm years Moldova builds about three dwellings per thousand inhabitants and has no cheap tier — is not explained by the lag. Here there is a measured benchmark: in the World Bank's last Doing Business ranking (2020) Thailand stood 34th of 190 for dealing with construction permits, at 113 days for the whole procedure; Moldova, from the assistant's memory and unchecked against the primary source, stood in the middle of the second hundred with a procedure roughly twice as long. One indicator tells the causes apart: the number of building permits issued and the time to obtain them, year by year. If, with the price up 60 %, many permits are applied for while few are issued and slowly — it is a barrier; if few are applied for — it is money, people and the lag.
Testing the indicator. The architect objected: "if the price rose 60 %, was it net profit that rose, or did materials and cement rise as much? There was no force majeure in Moldova in these years. You find no direct sources because they are hidden: that is corruption" — and asked for a check on land: can one build on agricultural land in Thailand, and how does that stand in Moldova. Three results.
Permits. The source is not hidden: the statistics office publishes the number of building permits issued. Residential buildings: 2,276 in 2024 (+1.1 % on 2023) and 2,161 in 2025 (−5.1 %). That is, in two years when the price per square metre rose 60 %, the number of permits did not rise at all and then declined — for a country of 2.4 million. And money in 2024–2025 was already cheap: the policy rate, raised above 20 % in 2022, had been cut several-fold. This is the indicator named above, and it answers in the architect's favour: the lag and the price of money explain the slump in completions of 2024, but not why supply failed to respond even with applications to a one-and-a-half-fold price rise. The shock of 2022 was real — a war next door, inflation above 30 % — but by 2024–2025 it was over, and supply did not revive.
The margin. No construction price index for these years could be found; consumer prices rose by roughly 12–15 % in total over 2024–2025, and even with a margin for error building costs could not have risen 60 %. So most of the increase is not cement but margin and the price of land with a permit. The argument here is economic, not accusatory: in a market with free entry excess profit does not live long — it attracts new developers, who drive the price down. Excess profit that holds for two years while the number of permits falls is itself a sign of dear entry, whatever its nature.
Land. In Thailand, in the agricultural ("green") zone of a town plan a dwelling house and even a hotel may be built, a condominium may not; the permit is issued by the local administration and is required in areas under building control. In Moldova one may not build on agricultural land until its designation is changed, and it can be changed only after compensating "losses" to the state — under a 2025 government decision in all cases, regardless of purpose or form of ownership, at a tariff approved annually. Conversion of land is a separate procedure with its own signatures, and it is compulsory even for a private house on one's own plot. This is not a check of a result (whether the house is sound) but a check of process in its pure form, and by the filter of §5 it is the first candidate for a "repeal" line or for reduction to a notification.
The architect's formula: "so Thailand simply torments its people less. You can build a private house anywhere, but to cut down the maize on three hundred square metres, whose whole harvest is worth 100 dollars, and build a house — in Moldova you cannot." The whole asymmetry is in this image: the state protects a hundred dollars' worth of harvest from its owner at the price of a house he will not build, and charges for lifting the protection a fee it calls "compensation of losses". Nobody has any losses: the maize belongs to the same person as the future house. The exact limits, so that the formula cannot be disputed: "anywhere" in Thailand means on one's own land with a full title, including the agricultural zone; not on land granted under the agrarian reform for cultivation only, nor in forests and parks, and in areas under building control a permit from the local administration is needed. The difference between the countries is not that one has rules and the other does not, but in what they protect: in one — the neighbours and the forest, in the other — the procedure.
The same "x2", only without a line in the law. The architect: "if the cunning state brazenly made a car x2 and you found all the figures, why do you assume it is different with flats?" The reproach is fair: the assistant's caution was asymmetric. With a car the mark-up is written as one line in the excise table and can be computed to the leu (§6e). With a flat there is no such line — but that is a difference in the method of collection, not in substance, and figures of the right order can be found just the same.
| Component of the price of a square metre in Chișinău | Order |
|---|---|
| Sale price | about 1,720 euros |
| Turnkey construction, without land and connections | 425–715 euros (a Moldovan construction publication's estimate for 2026; the first version of this row held the assistant's guess of 600–800) |
| Land per metre of flat | on the order of 125–250 euros: an are (100 m²) in Chișinău costs 50–70 thousand euros, over 100 thousand in the best places, and about eight thousand metres are built on twenty ares |
| The remainder — permits, connections, margin | on the order of 750–1,150 euros, that is, 45–65 % of the price |
The same factor of two results, and at the lower bound of cost more than that: a flat costs two to four times what it costs to build. And the main figure here is the price of land. A hundred thousand euros per are in a country people leave, with arable land lying idle around the city, is not the price of land but the price of the status "for building". The neighbouring plot under maize costs tens of times less only because one may not build on it until its designation is changed. The scarcity of land is created by the norm, and the difference in price is the capitalised value of the permit.
Construction cost: a comparison with Thailand. The architect supposed that the cost of building in the two countries is comparable; the check confirms this, and the table row above has been corrected from a source. The first version of the comparison used Bangkok prices; the architect objected: "do not take Bangkok prices, it is a world capital, everything is expensive there; take Pattaya." The objection is accepted: it is fairer to compare Chișinău with an ordinary Thai city, and Pattaya is moreover a resort with foreign demand, so the comparison is not tilted in Thailand's favour.
| Chișinău | Pattaya | |
|---|---|---|
| Turnkey construction cost, without land and connections | 425–715 euros per m² (a Moldovan construction publication's estimate for 2026) | 25–30 thousand baht per m², roughly 650–790 euros (a Thailand-wide estimate for an ordinary condominium or house) |
| Sale price, median of listings | a flat — 1,720 euros per m² | a townhouse about 22 thousand baht per m² (roughly 580 euros); a house about 37 thousand (roughly 980 euros); a flat on average across the city about 58 thousand (roughly 1,500 euros), a modern one by the sea 75–95 thousand |
| How many times the price exceeds the cost | 2.4–4 times | a townhouse roughly at the cost of construction; a house 1.2–1.5 times; an average flat about twice; by the sea two to three times |
Building costs about the same in both countries — concrete, rebar and labour cost much alike. Everything else diverges. In a resort city where people from all over the world come to buy property, a family townhouse sells at the cost of construction, a house at a mark-up of a third, and only a flat with a sea view reaches the multiple that in Chișinău is the city average. In absolute figures a metre of townhouse in Pattaya costs a third of a metre of a Chișinău flat, and a metre of house almost half. For Bangkok the picture is the same shifted upward: a townhouse about 31 thousand baht per m², a house about 58 thousand, and only a flat in the business centre of a megacity of eleven million (about 135 thousand) carries a mark-up of three to four times, like the average Chișinău metre. This is the measured price of the status "for building". Caveats: the Moldovan cost figure refers to a private house, a multi-storey block costs more, and for it the multiple is closer to two and a half; the Thai prices are medians of listings, not of transactions; mass-built townhouses are built more cheaply than the general estimate, so "at the cost of construction" means a small but positive mark-up.
What the gap consists of. The architect's question: "why is the same housing, of the same concrete and steel, with the same windows, cheaper in Thailand? What are the suppositions about hidden costs?" Since construction costs the same, the whole gap — on the order of a thousand euros per metre — sits not in the building but around it. Below are hypotheses, from the weightiest to the less weighty; these are the assistant's suppositions, not measurements, and for each a way of checking is named.
| Hidden cost | How it gets into the price of a metre | How to check |
|---|---|---|
| Time | The developer's money is frozen for as long as land conversion, planning documents, the permit, connections and acceptance take. At a cost of capital of 10–15 % a year, each extra year adds 10–15 % to the price of land and all early outlays. In Thailand a building permit takes 113 days (§6c above) | How many months pass from buying the plot to the first concrete, and from the last concrete to the act of acceptance |
| The risk of total loss | A project may be stopped after the money is sunk: the permit revoked, a court challenge, a moratorium, new rules under a new administration. Whoever risks losing everything builds in not a 10–15 % margin, as in a predictable country, but 30–40 % | The share of halted and frozen projects; developers' margins in their accounts |
| The status of land | One may build not where there is land but where there is status; scarcity of status yields 50–100 thousand euros per are | The price difference between adjacent plots with different designations |
| Few players | Only the large and the "insiders" can bear the time and the risk; with three to five developers per city nobody drives the price down | The share of the five largest developers in completions |
| Instalments built into the price | Mortgages are weak, and the developer himself acts as a bank: "five-year instalments at no interest" means the interest already sits in the price per metre | The discount for paying at once: that is the price of the embedded credit |
| Connections | Water, sewage, electricity and heat often have to be brought by the developer himself, who builds networks and substations and hands them to the monopolist | The cost of technical conditions per metre of housing |
| Unofficial payments | At every gate, and dearest at the last (see above) | Not measurable directly; indirectly — by the difference between statutory and actual timelines |
| Housing as a safe | Where banks and the currency are not trusted, remittances from abroad are parked in flats; demand exceeds the need for housing | The share of flats standing for years without finishing or occupants |
What does not belong here, though it comes to mind first: climate. Insulation, heating and a deep foundation really are needed in Moldova and not in Thailand — but they already sit in the construction cost, which, as shown above, is comparable.
The first six rows have one thing in common: they are the price of unpredictability and the price of waiting. Concrete costs the same in the two countries; what costs differently is time and the confidence that what is built may be sold. And almost all of it is not a tax that somebody received but a pure loss: the year a building waits for a signature goes to nobody. So subtraction here yields more than mere redistribution: cutting the time from three years to four months lowers the price per metre while taking nothing from the budget or from an honest developer. The last row — housing as a safe — is cured not by building but by trust in money, and that is already a slow effect (§6b).
The difference from cars lies in who receives the mark-up, and it is not in the flats' favour. The excise on a car at least reaches the budget. The mark-up on a flat for the most part does not: it settles with whoever already holds a plot with the right status and with whoever grants that status. That is why there are no direct figures — not because they are hidden in the statistics, but because nobody collects this tax officially: it is levied through the price. By this section's classification it is a levy, not a tax (§6e), and thereby a candidate for subtraction: abolishing compulsory conversion of designation for a private house on one's own land costs the budget almost nothing — it loses only the fee for "compensation of losses".
The last gate. The architect: "we have what we guard. And you have not even reached the commissioning commission yet, and the people there are rich." The wealth of commission members was not checked against open sources; it is the architect's observation. But the construction can be checked, and it explains why the post is lucrative. Under the acceptance procedure in force (a government decision of 1996) the investor presents to the commission the opinions of the fire service, the sanitary service, the environmental authority and "other control bodies"; without the act of acceptance the building does not exist legally — it cannot be registered, sold or occupied. The economics of this spot is known as the hold-up problem: the later the gate, the dearer the signature. At the stage of land conversion a developer can still walk away; with a finished building, loans drawn and flats pre-sold, there is nowhere to go, and every day of waiting costs money. So the last signature in the chain is the most expensive, and a chain of several independent signatures is several last signatures.
Here the filter of §5 gives an answer other than "repeal", and this is worth noting so that the section is not read as a call to tear everything down. Acceptance of a finished building is a check of a result: will the floor slab hold, are there exits in a fire. It should stay. What should be subtracted is something else: discretion and multiplicity. One inspection against a published checklist instead of opinions from several agencies; a deadline after which silence counts as consent; refusal only in writing, citing an item of the checklist; liability for hidden defects on the insurer and the designer, who have something to lose, rather than on the signature of an official who has nothing to lose. Each of these measures is known and works somewhere; none is introduced where those who decide elections do not care how buildings are commissioned.
What remains. The number of applications submitted is unknown, so "few are issued" and "few are requested" are not formally separated; the shortage of builders due to emigration also works in the same direction. But the claim "the sources are hidden" was not confirmed — the data are open, and they speak against the system in force by themselves. 🟡
What these figures do not prove. Why exactly completions fell in Moldova has not been established here from open sources; market participants name permitting procedures, but other forces also acted on the price — refugees from Ukraine, diaspora money, subsidised mortgages, inflation. So housing is recorded not as a proven levy but as a candidate for testing by the filter of §5: how many months and signatures a building permit takes compared with ten countries, and which of them check a result (structural soundness, fire safety) and which a process. 🟡
6d. Three thirty-second spots¶
The architect on the maize formula: "oh, how well this would play at an election, on television." It would, and it is clear why: it has an object that can be shown, a sum that can be compared, and not one word that requires an education. The language of subtraction is televisual in general — unlike the language of construction, where one has to promise what cannot be seen. The rule of a spot: one object, two figures, one question, one word at the end.
Maize. Shot: three hundred square metres of maize, behind it an empty foundation. Voice: "This is Ion's maize. The whole harvest is a hundred dollars. This is Ion's house. It does not exist. To cut down his own maize on his own land and build his own house, Ion must pay the state for losses. Whose losses? The maize is Ion's. The land is Ion's. The house is Ion's. Repeal."
The driving school. Shot: a young man reverses confidently into a parking space; the instructor in the next seat looks at his phone. Voice: "Andrei has been driving since he was sixteen. He will pass the test tomorrow. But first — thirty lessons and ten thousand lei. England has none of this, and fewer people die on its roads than on ours. The test — keep. The levy — repeal."
The cash register. Shot: a small roadside shop; on the counter a box with a new cash register. Voice: "Maria's shop has one shop assistant — herself. The device, its servicing and its connection to state monitoring are at her expense. This year she was told to replace the model — again at her expense. In Thailand a shop with that turnover needs no register at all, and the country is no poorer. Repeal."
The last frame in all three is the same: "Why has nobody repealed this? Because those who decide elections do not care. AB-EXIT changes who decides." Thus a spot about a levy becomes a spot about the protocol and answers the question "what do I get besides the dividend" without a single figure about GDP.
Two cautions. Every figure in a spot must withstand checking — one refuted sum ruins the whole series, so sums are taken from open price lists and statistics, while names and scenes are composite and said to be so. And a spot must not be built on accusing particular people: the argument "the norm checks a process, not a result" is stronger than "they steal", because the first is verifiable, while the second gives the opponent grounds for a lawsuit and for pity. 🟡
6e. A levy and a tax are not the same thing: importing a car¶
The architect's question: "I earned money honestly and paid my taxes. I decided to buy a car. Moldova has no car plants. Why must I, with my own money, buy a car from the United States and pay a tax? What is the logic of the tax — apart from all the poor being glad that this rich man with a car paid into the budget?" How it is arranged: customs duty on cars from most countries is 0 %; one pays an excise depending on engine size and the car's age (three bands: up to 10 years, 11–15, over 15; hybrids get a discount, electric cars are exempt), a customs fee of 0.4 % and a luxury tax on cars dearer than 600 thousand lei; a 20 % VAT on import is paid on commercial imports; for individuals its introduction, planned for 2026, has been postponed (see the calculation below).
A correction of fact and a worked example. The first version of this point said that from 2026 a 20 % VAT on import is introduced. A check against a current customs-clearance guide (April 2026): the introduction of VAT for individuals importing a car for themselves was postponed in December 2025; VAT is paid on commercial imports and on lorries. An individual pays the excise — engine capacity in cm³ multiplied by a rate in lei that depends on age and engine type — and a fee of 0.4 % of the customs value. The calculation for a 2018 Ford Fusion (age 8 years, rates from the excise table):
| Engine | Capacity | Rate, lei per cm³ | Excise | Roughly in euros |
|---|---|---|---|---|
| 1.5 petrol | 1,499 cm³ | 15.61 | about 23,400 lei | about 1,200 |
| 2.0 petrol | 1,999 cm³ | 23.68 | about 47,300 lei | about 2,400 |
| 2.5 petrol | 2,488 cm³ | 38.49 | about 95,800 lei | about 4,800 |
| 2.0 hybrid | 1,999 cm³ | 17.76 | about 35,500 lei | about 1,800 |
| 2.0 plug-in hybrid | 1,999 cm³ | 11.84 | about 23,700 lei | about 1,200 |
On top of this comes a fee of about 600 lei for a car worth, with shipping, some 8–9 thousand dollars. The most common version in the United States is the 2.5; for it the total is about 96 thousand lei, roughly 5.5 thousand dollars. How much that is of the car's price in the United States: in a private-party sale such a Fusion is valued at 8.5–10.9 thousand dollars — the tax is 50–65 % of the price; as a dealer trade-in, 4.4–5.7 thousand — the tax equals the price of the car or exceeds it; at a salvage auction, where these cars usually come from, sales at 850–3,700 dollars occur — the tax is one and a half to six times the price. The tax does not depend on the price at all: a wrecked Fusion and a perfect one pay the same, while half a litre of engine doubles the sum.
Two questions from the architect about this figure. "Is this not a customs duty, then?" Legally it is an excise: the duty is zero because its ceiling is bound by trade agreements, whereas an excise counts as an internal tax that a country may levy on a good regardless of origin. But cars are not produced in Moldova, so the "internal tax" is paid only at the border and only on what is imported — it is a duty under another name, and the name is the whole point of it. "And why is it tied to the engine and not to the wheel diameter or the width of the exhaust pipe?" Because engine capacity is written in the car's documents and cannot be understated, unlike the price, which is understated all the time; this is again a choice of what is easier to collect, not of what is fair. Thirty years ago capacity was a tolerable sign of an expensive and thirsty car. Today a 1.5-litre turbo engine is more powerful and more expensive than a simple naturally aspirated 2.5, yet pays a quarter as much: the tax punishes cheap reliable technology and rewards expensive complex technology. One could with equal justification charge by wheel diameter. This is a tax not on wealth and not on harm but on cubic centimetres.
Such a tax can in principle rest on five logics, and they are worth checking one by one.
| Logic | What it asserts | Does it hold in Moldova |
|---|---|---|
| Protecting the domestic producer | Dear imports let local plants grow | No: there are no plants, and the duty is honestly zero. Thailand, by contrast, taxes car imports at one of the highest tariffs in the world — but it has a car industry it protects |
| A tax on consumption | VAT is paid on everything from bread to a car; it is not "a second tax on the same money" but a different tax — on spending, not on income | Yes, if it is ordinary VAT at the general rate |
| A charge for harm | A car wears the roads, pollutes the air, creates congestion | Partly: the harm depends on mileage, and mileage is already charged through the excise in the price of fuel; the age bands make environmental sense but hit the poor — an old car is bought by someone who has no money for a new one |
| A drain of currency | Imports "take money out of the country" | No: eighteenth-century mercantilism; a person earned the money and is free to exchange it for what he likes |
| A tax on visible wealth | Where incomes are in the shadow and arrive as remittances, the state taxes what cannot be hidden: a car, a house, an import | This is the real reason, and it is the one the architect named |
A correction to the row on the consumption tax. The architect: "and what, is VAT everywhere?" No, and the table row presented it as a law of nature. The United States has no VAT at all — it has state sales taxes, from zero (in five states) to roughly ten per cent; there is none in Hong Kong, Kuwait or Qatar, and in the Emirates and Saudi Arabia it appeared only in 2018 at a rate of 5 %. Where it exists, the rate differs fourfold: 7 % in Thailand, about 8 % in Switzerland, 9 % in Singapore, 10 % in Japan, 20 % in Moldova, 27 % in Hungary. So only the narrow statement is true: within a country with VAT a car is taxed the same as bread, and in that sense the tax is neutral. But the very fact of VAT and its height are a decision like any other, and twenty per cent against seven is not nature but a choice someone answers for. This does not make VAT a levy — it goes to the budget — but it returns it from "a given" to the list of parameters that those who stay to vote have reason to ask about.
"A tax that requires no thinking." The architect: "VAT is a tax on everyone that requires no thinking. AB-EXIT switches on the work of government. With lower spending, perhaps it can be abolished." The first half is exact. VAT collects itself: everyone pays it on every purchase without seeing the sum, it cannot be hidden, and a government that lives on it needs neither to know citizens' incomes nor to negotiate with them. In Moldova it is not one tax among others but the foundation of the budget: in 2024 VAT provided more than half of state budget revenue — about 51 % in the draft, and roughly 30 billion lei out of 60 over eleven months. A state that lives half on a tax requiring no effort has no reason to learn to collect the rest. And this tax has its retinue: it is for its sake that a shopkeeper is given a cash register connected to online monitoring (§1); a country with a threshold of 1.8 million baht relieves all petty trade of that retinue at a stroke.
The second half requires arithmetic. The repository's cautious estimate is a release of 20–30 % of what is lost today (019.59). If that is taken as a share of state budget spending, then with VAT at half of revenue it suffices for roughly half of VAT: a rate of 20 % becomes 10 %, not zero. Full abolition would require either twice the spending cut the repository considers attainable, or replacement by another tax — and tax scholarship in fact regards a broad consumption tax at a low rate as one of the least harmful; high taxes on labour, which drive wages into the shadow, are worse. So the honest formula for a campaign is not "we will abolish VAT" but "a government that has started working returns half of VAT to people and raises the threshold below which petty trade is left alone entirely". Ten per cent instead of twenty is a lower price for everything a person buys, and a sum that needs no explaining.
A caveat: the calculation is crude — the VAT share is taken for the state budget, not the whole national public budget, where with the social funds it is lower, and cuttable spending is distributed differently. This is a parameter for economists (001b.12b); what matters here is the order: not zero and not pennies, but about half. 🟡
The car and the loaf. To the argument "tax scholarship regards a broad consumption tax as the least harmful" the architect objected: "that is what they all counted as science before AB-EXIT. It is a tax for a lazy government — to work less and count worse. VAT on a car for 10,000 brought from the United States — where is the value added here? And VAT on a loaf of bread baked from flour from the neighbouring field — compare." The comparison is worth making literally.
| A 10,000-dollar car from the United States | A loaf of bread from local flour | |
|---|---|---|
| Where the value was added | Wholly abroad; nothing was added in the country | Wholly in the country: the field, the mill, the bakery, the shop |
| How much is taken | On a commercial import — 20 % at once, 2,000 dollars; from an individual, instead of VAT, an excise on cubic centimetres, about 96 thousand lei for a Fusion 2.5 | At the reduced rate on bread; fractions of a leu per loaf |
| Who collects and how | One customs officer at one barrier; do not pay and the car is not released | Four taxpayers along the chain, each with invoices, offsets, a cash register and inspections |
| What collection costs | Almost nothing | An accountant, a device, online monitoring and an inspector at every link |
This shows that the tax's name deceives. "Value added tax" is a method of collection: a tax on consumption which inside the country is collected in parts along the chain. At the border there is no chain: the whole sum is taken at once from whoever brought the goods, simply because he crossed the barrier. And for a lazy government this is the best part of the tax. By the reports of Moldova's customs service, it collects on the order of 28 billion lei in nine months against state budget revenue of about 60 billion in eleven — that is, by order of magnitude, about half of all revenue comes from the border (VAT, excises and duties together; the source does not state the year of the customs figure, so this is an estimate, not a calculation). The result is a closed circle worth naming plainly: people leave to earn, send money home, families buy imported goods with it, and the state takes a fifth at the barrier. The budget of a country people leave consists in large part of a tax on the labour of those who left, and collecting it requires not a single working institution.
On the science. The correction is accepted by half, and it is worth saying which half. The argument "VAT is the least harmful" is derived on the assumption that the quality of government and the volume of spending are given from outside; it answers the question "how to collect this sum with least damage", not "whether it should be collected". But the other half of the same science — the public choice school — says exactly what the architect says: a tax the payer does not see and which collects itself lets the state grow unasked; in Buchanan this is "fiscal illusion", in the VAT literature it is the question "is it a money machine", and it is for this very reason that the United States still has no VAT. The assistant cited the first half and was silent about the second.
The goose and the hissing. The architect on the formula "collect with the least damage": "least noticeably and covertly. It is nowhere except on the receipt. Every direct tax has people who suffer from it. This is a way to smear injustice as thinly as possible over everyone." This is an old admission of tax practice itself: Colbert, minister of Louis XIV, is credited with the formula that the art of taxation is to pluck the goose so as to get the most feathers with the least hissing. A direct tax has a person who received a bill, knows the sum and knows to whom he pays it; he hisses. VAT has no such person: the sum is dissolved in the price, the payment is split into thousands of purchases, and there is nobody to be indignant and nothing to be indignant at.
It is the same mechanism as in §2 taken to its limit: a levy lives when the payers are dispersed, and VAT disperses the payers completely — everyone pays, a little, always, and without seeing it. And here too is the answer to the question with which this point began. "The poor are glad the rich man paid for his car" — but relative to income the main payer of VAT is precisely the poor person: he spends everything he receives and gives up a fifth on every purchase, whereas whoever saves pays nothing on what is saved. The gladness is paid for out of his own pocket, and paid unnoticed.
A clarification, so as not to argue about words. In economics "least harmful" means "distorts behaviour least": VAT does not blunt the wish to work and invest the way a high tax on labour does, and that property is real. But the political reason governments love it is different — the one the architect named: it hisses least. Both properties are true at once, and the second matters more for the question the protocol asks: not "how to collect more quietly" but "who will ask, and when, what the collected money was spent on". Hence the practical consequence above — visibility: the sum of tax as a separate line on every receipt and an annual total for each person, beside the sum of the dividend. A goose that sees how many feathers were taken begins to hiss — and that is feedback switched on.
What follows and what does not. It follows: ease of collection is a virtue for the collector, not for the payer, and a protocol that "switches on the work of government" should make the tax visible — so that on the receipt and at customs a person sees the sum he pays as clearly as the sum of the dividend. It does not follow: the arithmetic above stands — half the budget cannot be removed with one line, and the order of the attainable is roughly halving the rate and a high threshold for petty trade, not zero. 🟡
The wolf and the lamb. The architect's summing-up: "so I pay double for importing a car for myself personally only because the state obliged me, and that is all? And the state took its mandate from a majority that has no car and does not know what it is. Everything you write about the car resembles 'you are guilty merely because I want to eat'." Yes, and the fable describes the point more exactly than the table. Krylov's wolf does not say "I am hungry" at once — he goes through pretexts: you muddy the water, you were rude last year, if not you then your brother. The table of five logics above is the same pretexts in turn: we protect the producer (there is none), we save currency (it is yours), we recover harm (already charged in the price of fuel), we tax luxury (a wreck from an auction pays the same as a new car). When the pretexts run out, the last line of the fable remains. The honest wording of this tax is: "the state needs money, and it is easiest to take it from you." It is called that nowhere precisely because under that name nobody could vote for it. The lamb in the fable loses not the argument — in the argument he is right on every count — but because he has no weight. The forecast of weight is in §6g.
The last line is not stupidity but a diagnosis. A tax on importing a car in a poor country is the state's admission that it cannot collect income tax honestly, so it collects at the border, where there is one barrier and collection is cheap. A person who has already paid income tax pays a second time for those who did not. "The poor are glad" is a consequence, not a cause: for the majority without a car the tax costs nothing, and a politician who lowers it gains no votes — the same three conditions of §2.
But the boundary with a levy is one of principle here, and it must not be blurred. A levy goes to an intermediary, and repealing it costs the budget nothing (§4). A tax goes to the budget — the very one from which the dividend is paid — and repealing it requires either a replacement or a cut in spending. So importing a car does not belong on the list of "a hundred lines of 'repeal'": it is a question of the structure of taxes, that is, a parameter for economists (001b.12b), not reform by subtraction. What the protocol changes here: the composition of those who vote shifts towards those who pay taxes, and the question "why this tax and not another" for the first time gets a voter to whom the answer is not indifferent. The sensible direction of such an answer is known from tax scholarship: a broad base and a low rate instead of a narrow base and a high one, and a charge for harm through fuel, that is, for mileage, not for the right to own. 🟡
6f. Returning taxes: the dividend of good governance¶
Two questions from the architect. The first: "perhaps the United States is the most successful and cleverest country precisely because it has no 20 % tax on everything?" The second: "if after AB-EXIT clever rulers have improved governance, can they lower or return taxes? In business that is called dividends."
The first is half proven. The absence of VAT in the United States is part of a more general fact: the state there takes about 27–28 % of GDP against roughly 34 % on average in developed countries and about 40 % in Europe, and hiding a tax rise in prices is harder there. But the data do not bear the causal link "no VAT, hence success": among the rich there are countries with no VAT or a low one (the United States, Hong Kong, Switzerland, Singapore) and countries with VAT at 25 % (Denmark, Sweden). What tells them apart is something else — what the state does with what it collects. A high tax under good governance people tolerate and even approve; a high tax under bad governance is the worst of both worlds, and that is Moldova's case: a European rate with a non-European quality of spending. So the exact formula is not "VAT makes a country poor" but "an invisible tax lets bad governance avoid answering for the price it costs".
The second — yes, and it already exists. Returning what was collected beyond need is working practice, not fantasy.
| Where | How it works |
|---|---|
| Colorado, USA | A constitutional rule of 1992 (TABOR): state revenue above a cap that grows with population and inflation is returned to taxpayers by formula; for fiscal year 2023–24 — 1.7 billion dollars, about 326 dollars per person on average |
| Alaska | The Permanent Fund Dividend to every resident annually (014) |
| Hong Kong, Macau, Singapore | One-off and annual payments to citizens out of surpluses; in Singapore some of these payments were actually called "growth dividends" (from the assistant's memory, unchecked against the primary source) |
For the protocol this is a second channel, and it is important not to confuse it with the first. The dividend D is paid to whoever declined to vote, is computed by formula from the median and does not depend on the government's will (001b.12b). A tax return goes to all taxpayers and depends precisely on the government's work: saved — returned. Together they give what no state has today: D shows whether it is worth taking part, the return shows whether governance was good. It is a third number beside the two sensors (040b.2b), and the most direct one: a company that has earned pays its shareholders; a company that does not pay explains why.
Lower the rate or return the money? By the logic of the previous point — return it. A rate cut dissolves in prices as unnoticed as VAT itself; a return arrives as a sum in the account that a person sees, remembers and compares with last year's. The rate is worth cutting once the return has become stable — then the cut locks in what has been achieved.
Two conditions without which a return turns into the same pre-election buckwheat. It must follow a rule, not a decision: the formula of the cap is written in advance, as in Colorado, and the government has no button "hand out before the election". And only what was really saved is returned: a return out of debt or out of underfunded hospitals is not a dividend but a sale of assets. The honest remainder: Colorado's experience also shows the other side — legislators constantly look for ways to redirect the surplus, and in downturn years a rigid cap hinders restoring spending; the rule is needed, but it should be written by economists, not by a slogan. 🟡
6g. A forecast: how many of those who stay are touched by the tax on a car¶
The architect's question: what percentage of voters after the protocol will have a car, realistically want one in the future, or know the price of customs clearance through relatives — "perhaps this will no longer be a minority". Below is a model, not a measurement; the assumptions are named so that they can be challenged one by one.
The starting point. Moldova has about 745 thousand registered passenger cars (statistics for the end of 2022); those actually alive are estimated at about 600 thousand. There are roughly 1.9 million adults and about a million households. About 40–50 % of households have a car, and about half of adults live in them, because families with a car are larger. No direct figure from the household budget survey could be found; this is an estimate. Among those who vote today the share is lower — roughly 40–45 %, because the elderly and rural residents turn out better. The architect's formula "the majority that votes has no car and no licence" is true of today, but the margin is small.
Who leaves. The sum instead of the vote is more often taken by the poor, the elderly, people with a short horizon — that is, more often by those who have no car. Two variants at an overall exit share of about 60 %:
| Stratification | Leaving among those without a car | Leaving among those with a car | Share of "a car in the family" among those who stay |
|---|---|---|---|
| Strong | 75 % | 45 % | about 69 % |
| Weak | 65 % | 55 % | about 56 % |
Who is personally affected. To the owners are added those who have no car but intend to buy one, are getting a licence, or know the price of customs clearance through their children and relatives. If that is half of those who stay without a car — and for the young and employed that is a cautious estimate — the result is 75–85 % of those who stay.
The conclusion. On everything connected with a car — the import excise, the compulsory driving school, registration fees — a minority very probably becomes a majority of two thirds or three quarters. The third condition for a levy's survival (§2: "the majority that decides elections does not care") ceases to hold for this group of norms.
This also closes the question of visibility left open in §6e. The architect objected to the thesis "VAT is visible nowhere": "when importing a car I see it clearly and distinctly, and I see it on the receipt". The correction is accepted: a person sees the sum at customs, gets angry — and nothing changes, because those who see it are few. Visibility in itself is not feedback; it becomes feedback when the one who sees acquires weight at elections. An annual statement with a total is needed against a tax smeared over a thousand receipts; against a tax paid by few and in large sums only a change in the composition of those who vote works.
Caveats. The main assumption — that car owners take the sum less often — is plausible (a car in Moldova is a good proxy for income and planning horizon) but unmeasured; it is a question for a survey or a pilot. Present owners have a double interest: cheap imports make the next car cheaper but lower the resale price of the present one; they will not vote as a single bloc. The diaspora was not included; it has more cars, and its relatives pay the Moldovan customs clearance, so by staying in the vote it strengthens the same majority. 🟡
6h. The common denominator is the state. Georgia: subtraction is possible¶
The architect on the table of hidden costs (§6c): "what all of them have in common is that all this depends on the state, not on builders or buyers; Thailand and Moldova have different authorities and different effectiveness." This is more exact than the assistant's formula "the price of unpredictability and waiting": time, risk, the status of land, the number of players, connections and gates — none of them is created by the builder or the buyer. The same bricklayers build from the same concrete in the two countries alike; the only difference is what happens before the first concrete and after the last. The same holds for the driving school, the cash register and the car: in every example of this section the object costs the same, and the difference in price is created by paper.
A caveat, so as not to idealise. Thailand is no model of democracy, and the lightness of its state in everyday life is not the voter's merit: in large part it is a state that simply does not reach the street kitchen and sees no point in taxing it, and a country that competes for tourists and investors. A post-Soviet country inherited the opposite — dozens of supervisory agencies, each of which has to be fed. So the question for Moldova is not "how to become Thailand" but "can the inheritance be subtracted".
It can, and this has been tested on a country resembling Moldova. In Georgia in 2004–2006 Kakha Bendukidze's reforms cut the number of activities requiring licences and permits from 909 to 137; whole agencies and inspectorates were abolished. Before the reform only 207 of 484 construction projects in Tbilisi had a permit — the rest were built on bribes; afterwards the country rose in the ease-of-doing-business ranking from 112th to 17th place in six years. The reform was precisely subtraction: nearly eight hundred lines of "repeal", with no new institutions and almost no money.
What this experience teaches and what it lacked. It proves that subtraction is technically simple and gives a fast, manifold effect (§6b). But its engine was a single team on the wave of a revolution, not the voter: subtraction had no permanent client. So it stopped together with the team, and part of what was subtracted came back — in spring 2026 Georgia again raised the excise on imported cars older than six years, threefold. This is exactly what the protocol answers: it does not invent the reform — that is known and tested — but creates a permanent client for it, a composition of voters for whom levies cost their own money (§3). 🟡
A lock with a hole. The architect: "but Georgia rolled back because there was no referendum and no AB-EXIT." A check yields a refinement stronger than the thesis itself. One half — the referendum — Georgia did have: the Economic Liberty Act of 2011 wrote into the constitution that a new national tax may be introduced, or the upper rate of an existing one raised, only by referendum, and capped the spending of all budgets at 30 % of GDP, the deficit at 3 %, the debt at 60 %. But the norm carries a proviso in brackets: "except for excise tax". And the rollback of 2026 came precisely through those brackets — the excise on imported cars was raised. Hence two lessons for the protocol. First: a lock without a voter who needs it is bypassed through the first gap left open; writing a closed list of exceptions is useless if there is nobody to watch what gets into it. Second: Georgia had the referendum but not the other half — a change in the composition of those who vote; there is still nobody to call anyone to account for raising the excise on someone else's car. The protocol installs both halves: the rules change only by referendum (048g.6b), and the one who pays watches the gaps.
6i. A check against a real country¶
The general theses of this section were checked against the open data of one country — Moldova; by the "international first" rule the country material is set out in a separate section, 053b. In brief, what the check showed. Everything has been counted and published: the government has the figures (892 mayoralties, 87 % with fewer than three thousand inhabitants, about 30 % of a small mayoralty's budget spent on itself), reports on the reform have been appearing since 2015, and every village's budget is open on a portal with 1.3 million data points. And nothing has been done: the reform is possible only as a voluntary one, with the guarantee "nothing will close" and a payment for consent of 6.49 billion lei. In the country's smallest mayoralty two thirds of the budget goes on the office and 1 % on what will remain to the village, while eight more such offices operate within eight and a half kilometres. Amalgamation would give not "x10" but an eightfold cheaper office for the smallest village, threefold on administration and a fifth of the budget overall. The conclusion is general for any country: knowledge is not the bottleneck (§2); transparency supplied without demand changes nothing; what is missing is not data but a voter who needs them.
6j. Consumer protection as a hidden levy¶
The example came from the analysis of the jury court (048j) but belongs to this section. The architect, who sold electronics for many years, described an ordinary case: "my appliance broke, and it is logical that it should be repaired and returned to me. But I bought a new one and I want a new one, and besides I do not want to wait. All this is written a thousand times in the sales contract — but the state is almost always on the buyer's side, because there is no protocol and the populists are in charge, who could not care less about business and badly need contented, apathetic consumers of microwave ovens. Not everywhere, but almost everywhere: the state makes the seller hand over a new one at once and keep the old one himself. What does the seller do? He puts the risk into the price. The populist rides high, and thousands of other buyers have paid for it all."
This is the same anatomy as any levy (§1): the benefit is visible, concentrated and credited to a politician — a person got a new microwave "by law"; the cost is spread in a thin layer over the price for everyone and credited to nobody. There is one difference — the money goes not into a trough but from the quiet buyer to the noisy one; which is why this levy is especially durable: it does not even have a recipient one could name. Who pays is visible in the figures: in the USA fraudulent and abusive returns in 2024 came to 103 billion dollars, about two per cent of all retail revenue (Appriss Retail and Deloitte); with retail margins of a few per cent that is a third to a half of profit, and the seller builds it into the price. The one who pays is the person who honestly has the broken thing repaired and waits — for the one who demands a new one at once.
The assistant tried to mark a boundary: buyer protection as such is not populism — the seller knows more about the goods, and without rules on defects the market deteriorates. The architect did not accept this either: "reviews by repairmen and buyers are better than rules, which can be got round and are new every day. Any washing-machine repairman will tell you over the phone for fifty dollars what not to buy — you need not even call him out; repairmen's numbers are on every site and forum. That is what brands are for: Samsung, not yesterday's newcomer 'megasuperluxeuromachine'. Tests of machines on YouTube with millions of views are better and more useful than state control."
The objection is right, and the assistant's argument is some thirty years out of date. The inequality of knowledge between seller and buyer was described by Akerlof in 1970 — and in the same paper he named what the market itself uses to cure it: guarantees, brands, chains, reputation. Since then something has been added that did not exist then: a repairman who sees hundreds of broken machines from the inside and answers the phone; thousands of reviews; teardowns and trials on video. A brand is a deposit: years and money invested in a name, lost through one bad batch; a fly-by-night has nothing to lose, and the name itself shows it. Today a buyer who has spent twenty minutes knows more about a washing machine than the salesman on the floor. A state standard, meanwhile, checks paper, is updated once a decade and is circumvented the day it is adopted. It is the same conclusion already recorded for food and driving (015b.6c): not "the market instead of the state" but "free entry plus a check of the result instead of permits and paper checks".
What remains of the role of rules is two things, and neither is about consumer protection as a separate branch. First: what was promised must be delivered — a guarantee given by a brand or a seller must be worth what is written on it; that needs not a special code but a fast court (048j). Second: danger to life — fire, electric shock — is a matter of checking the result, like the samples of cooked food in Bangkok, not of checking the process. What turns today's protection into a levy is the structure of penalties — a fine in the claimant's favour, an uncapped penalty, expert examination at the seller's expense — under which checking a claim does not pay the seller whatever the outcome, and the winner is not whoever is right but whoever is more brazen.
After the protocol the addressee changes. The apathetic consumer of microwave ovens is precisely the one who takes the sum and exits; his gratitude at the polls is no longer worth anything. Those who remain are those who count: they see that the "free" replacement was paid for by themselves through the price, and a rule written for the sake of a picture loses its sponsor. Buyer protection need not be abolished — it is enough to remove the lever of extortion from it; and a fast, cheap way to establish whether there was a defect is provided by the jury court (048j). 🟡
7. Weak point of the section¶
A delay of one electoral cycle does remain: repeal comes through the first election after adoption, not on the day of adoption. Under autocracy the tangle and the levies benefit the very top as an instrument of control (015b.6c), and those who stay will not repeal them until they change the top. The estimate "hundreds of millions of lei" is an order, not a calculation: the number of licences issued was not found in open sources. The main assumption — that those who stay really do pay levies more often than those who leave — is plausible but unmeasured. An indicator for a pilot is simple and cheap: whether, after the first cycle, candidates' programmes begin to contain promises to repeal specific levies that were not there before, and how many lines are repealed per cycle. 🟡
Related: 015b.6b–6c (the thermostat, the low base, the checking of the examples) · §19 (the cost of bad governance) · 019b (verifiable promises) · 018.40–41 (the pizzeria test) · 015c.7b (organised groups) · 033c.9c (the order of the sums) · 040b.2b (two sensors) · 053 (Moldova)