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.

49. The SMS to the Citizen¶

Chapter: 09 File version: v1 Date: 2026-06-11 Source: v6.53 §10, §11, §12, §13, §14

An early-draft layer (note of 01.10.2026 after audit 040m). The municipal statute in this chapter is the version from before the decisions of 30.09.2026: the bonus in the SMS, K = 1 % and the council's right to change the coefficient have been removed. The exact-answers sheet 1d and the charter 048m are in force.

The SMS below is the earlier version: the current protocol has no BONUS line, and the sum is a worked example.

┌──────────────────────────────────────┐

│ AB-EXIT Dividend 2028 │

│ │

│ BASE: $570 │

│ (W-2 median: $38,000 × 1.5 × 1 %) │

│ │

│ BONUS: $522 │

│ (the mayor saved $80M) │

│ │

│ TOTAL: $1,092 │

│ │

│ Trend: $890 → $1,092 (+22.7 %) │

│ ★★★★★ │

│ │

│ Cleveland: $640 (+12 %) ★★★★☆ │

│ Pittsburgh: $720 (+8 %) ★★★☆☆ │

│ Columbus: $810 (+15 %) ★★★★☆ │

│ DETROIT: $1,092 (+22.7 %) ★★★★★ │

│ │

│ Your W-2: $41,000 │

│ You are ABOVE the median ($38,000) ✓ │

│ │

│ Verify: abexit.gov/verify │

└──────────────────────────────────────┘

Stars	Condition
★☆☆☆☆	The dividend FELL by more than 10 %
★★☆☆☆	The dividend fell by 5–10 %
★★★☆☆	The dividend is stable (±5 %)
★★★★☆	The dividend grew by 5–15 %
★★★★★	The dividend grew by more than 15 %

ARTICLE X: CIVIC DIVIDEND CALCULATION

Section 1: FORMULA

The Civic Dividend (D) shall be calculated as:

D = M̄ × 1.5 × 0.01

where M̄ is the three-year CPI-adjusted average

of median W-2 wages for all ZIP codes

within the municipality.

Section 2: DATA SOURCE

Primary: Social Security Administration

W-2 wage data aggregated by ZIP code.

Amendment of 30.09.2026 (048k §3). Section 3 below reflects the earlier wording. The architect's decision: the size is set as a percentage of the median, determined by open debate and a referendum; there is no fixed 1 %; the council's power to raise K is removed; the percentage may be changed no more than once per cycle, by a simple majority.

Section 3: COEFFICIENT K

K = 1% (one percent), fixed in this charter.

City council may increase K up to 2%

if prior cycle savings exceed AB-EXIT cost

by factor of 3, with 2/3 council approval.

K above 2%: referendum with 2/3 majority.

K below 1%: referendum only.

Section 4: HOUSEHOLD COEFFICIENT

1.5 (one and one-half), fixed in this charter.

Reviewed once per decade following

decennial Census.

Section 5: TIMELINE

SSA shall provide W-2 median by ZIP code

within 120 days of January 31 filing deadline.

City shall publish dividend calculation

within 30 days of SSA publication.

Dividend paid no later than 1 day

before election.

Amendment of 30.09.2026 (048k §3). Section 6 is struck: the bonus for budget savings is removed, the protocol has one payment. The number and frequency of payments is a question for experts.

Section 6: EFFICIENCY BONUS

B = max(0, Budget_plan - Budget_fact) × 0.30 / N

Paid after independent GAAP audit,

within 6 months of election.

Section 7: CROSS-VERIFICATION

Annual comparison with IRS SOI, Census ACS,

and BLS QCEW data.

Divergence exceeding 10% triggers

independent audit by judicially appointed

statistician. Results published within 90 days.

Section 8: FOUR-AGENCY RULE

Dividend calculation continues as long as

ANY ONE of SSA, IRS, Census Bureau, or BLS

publishes income data for the municipality.

Section 9: ANTI-MANIPULATION

No local official shall have authority

to modify, adjust, or influence the

determination of W-2 median wages

or any variable in the dividend formula.

Variant A: The Census Bureau household median (ACS)¶

What it is: a survey of 3.5M households annually, a direct question about income.

Why it was considered:

Direct measurement of the household median (no coefficient 1.5 needed)
20 years of data history
Coverage of the poor (everyone is surveyed, including the non-working)

Why REJECTED:

CRITICAL DEFECT: AB-EXIT creates a $600+ incentive to inflate income in the Census survey. The punishment for lying in a Census survey = $0. In 230 years not one person has been convicted of lying to an enumerator. AB-EXIT tied to the Census WILL BREAK the Census as a data source — the system attacks itself (the protocol's autoimmune disease)
Accuracy ±5–15 % (a subjective survey, from memory)
The response rate is falling (97.5 % → 88 % over 20 years)
The error for cities <25,000 reaches ±20 %

Variant B: Budget tax revenue (General Fund Revenue)¶

What it is: a fixed percentage of the city's actually collected taxes.

Why it was considered:

Philosophical appeal: "the city = a corporation, taxes = revenue, the dividend = the shareholder's share"
Strict audit (Fitch, Moody's)
Impossible to hide billions of collected taxes

Why REJECTED:

CRITICAL DEFECT: the mayor CONTROLS tax rates. Raised taxes by 20 % → Revenue rose → the dividend rose → "well done, mayor". But in fact business left and people got poorer. Revenue REWARDS tax rises in the short term
Corruption increases Revenue (inflating the budget through excessive fees, fines, new charges)
Cutting taxes (good governance) is PUNISHED by a falling dividend
Double reading: Revenue growth = the economy is growing OR the mayor jacked up taxes? Impossible to tell without deep analysis

Variant C: The Federal Poverty Level (FPL)¶

What it is: a single figure from Washington ($15,060 in 2024), tied to a food basket.

Why it was considered:

Absolute protection from manipulation (a federal figure; the mayor does not control it)

Why REJECTED:

Identical for expensive San Francisco and cheap rural Arkansas — kills representativeness
10 % of the FPL = $1,500 — meaningless for LA and excessive for the heartland
Does not reflect the economy of the SPECIFIC city — tied to a national food basket

Variant D: The minimum wage¶

What it is: linkage to the local minimum hourly rate (for example, 100 hours of minimum wage).

Why it was considered:

Ideally reflects the cost of basic survival in the specific city
Understandable to every worker

Why REJECTED:

CRITICAL DEFECT: the minimum wage is set by LOCAL POLITICIANS. If the dividend is tied to the minimum wage, the elite will block to the death any rise in the minimum wage so as not to pay more dividends. AB-EXIT would inadvertently freeze the incomes of the poorest strata. A perverse incentive

Variant E: IRS AGI (Adjusted Gross Income) as PRIMARY¶

What it is: the median adjusted gross income from tax returns (Form 1040, line 11).

Why it was considered:

Punishment for lying: prison (a federal crime)
Coverage: 153M returns (100 % of filers)
Accuracy: ±1–3 % for legal incomes
Verification: the employer's W-2 is reconciled with the return

Why REJECTED in favour of the W-2:

A 2–3 YEAR lag (SOI publication). Unacceptable for operational dividend calculation
AGI ≠ real income: an entrepreneur with $200K of income may show an AGI of $94K after deductions
43 % file jointly (household), 57 % separately — we don't know who lives together
The poor (income <$13,850) don't file returns — invisible to the IRS
The W-2 solves all these problems: a 5-month lag, from the employer (not the worker), no deductions, no joint/separate problem

Variant F: Census ACS + BLS QCEW correction (hybrid)¶

What it is: the Census median as an annual calibration + the quarterly BLS trend for interpolation.

Why it was considered:

Census = a precise median once a year
BLS = a fast trend every quarter
The combination = fresh data with a precise base

Why REJECTED:

The Census remains PRIMARY → vulnerable to the inflation incentive (the same autoimmune problem)
BLS QCEW publishes the MEAN, not the median — a conversion coefficient (0.75) is needed, which DIFFERS across cities (from 0.65 to 0.82) — a source of error and of manipulation through the choice of coefficient
Double dependence on two agencies complicates the formula without a proportional gain in accuracy

Variant G: A composite index (ABEI — 4 metrics)¶

What it is: a weighted index of the income median (40 %) + employment (25 %) + population inflow (20 %) + safety (15 %).

Why it was considered:

A comprehensive assessment of the city's performance on 4 parameters
Each parameter covers an aspect the others miss

Why REJECTED:

"Your ABEI rose by 2.3 %" — the citizen: "What?! How much money?!" — INCOMPREHENSIBLE
Who sets the weights (40/25/20/15)? That is a POLITICAL decision. The mayor lobbies for the weight of the metric where he has good numbers — manipulation through WEIGHTS
The income median ALREADY CONTAINS all 4 metrics: employment growth → wages ↑ → median ↑; population inflow → median ↑; safety → people stay → median ↑. A composite index is redundant — the median = the integral indicator

Variant H: Coefficient 1.1 (a correction for pensioners + the self-employed)¶

What it is: an additional multiplier to account for incomes the W-2 does not see.

Why it was considered:

The W-2 does not count pensioners, the self-employed, investors — the coefficient compensates

Why REJECTED:

Pensions = a federal TRANSFER, not the city's income. Counting pensions = creating a perverse incentive (the mayor attracts pensioners instead of creating jobs)
The self-employed (1099) earn LESS than the median W-2 ($36K vs $45K) — adding them would LOWER the median, not raise it
The coefficient 1.5 (household) already contains an indirect correction
Three coefficients (1.1 × 1.5 × 1 %) = complicated. Two (1.5 × 1 %) = simple. Simplicity won without loss of accuracy

Variant I: K = 0.7 % (the original coefficient)¶

What it is: the originally proposed dividend coefficient.

Why REJECTED in favour of 1 %:

0.7 % = an ugly number, not computable in the head
"120,000 × 0.007 = ...er...840?" vs "120,000 × 1 % = 1,200. Full stop"
1 % = a brand. 0.7 % = a number
1 % is sufficient for filtering (the poor's risk premium = 400 %; even 1 % at subjective valuation beats vote-buying)
The budget balance at 1 % is better than at 0.7 % (+$94M vs +$78M for Moldova)

Variant J: K = 2 % (a strengthened coefficient)¶

What it is: a doubled coefficient for "countries under attack" (Moldova, Shor's vote-buying).

Why it was considered:

At K = 2 % the cost of buying a vote for Shor would rise 7 times
Stronger filtering of the electorate

Why REJECTED as the base:

K = 1 % is already sufficient to destroy vote-buying (the poor's subjective premium = 400 %, i.e. 1,200 safe lei = 6,000 subjective > 1,700 dirty from Shor)
~~K = 2 % = a loss of legitimacy (78 % take the dividend, only 22 % vote)~~ — inverted 28.09.2026 by the architect's position: 78 % taking is not a loss of legitimacy but the protocol's goal: those vote who came by reason, not by impulse; K is a tuning parameter for this result, not a dogma (055c §6.3, 033c.9c). The only remaining argument against K = 2 % is the budgetary one (next line)
K = 2 % = excessively expensive for the budget
Compromise: K = 1 % as the standard; raising to 2 % — a decision of the local council if savings exist (removed 30.09.2026: only a referendum changes the size)

Criterion	W-2 (chosen)	Census	IRS AGI	Revenue
The mayor controls it	NO	NO	NO	YES
AB-EXIT breaks the source	NO	YES (autoimmune)	NO	NO
Punishment for lying	5 years in prison, 97 % conviction	$0 (zero cases in 230 years)	5 years in prison	N/A
Protection from forgery	5 levels (triple copy + payroll software + worker-witnesses + prison + statistical impossibility)	None	2 levels (W-2 reconciliation + prison)	Audit (circumventable)
Lag	5 months	9 months	2–3 years	3 months
The citizen can verify	YES (own W-2)	NO	YES (return)	NO
Counts PRODUCTION	YES (only what is earned)	No (includes pensions, transfers)	Partly (AGI understated by deductions)	NO (counts rates, not the economy)
Understandable	YES	YES	YES	NO
Accuracy	±3 %	±5–15 %	±1–3 %	Precise but WRONG
Mass forgery	Impossible (9,000 simultaneous crimes)	Easy (5 % of respondents lie individually)	Hard but possible	Not applicable

The W-2 = the only source that AB-EXIT cannot break, that the citizen can verify, and that measures the city's REAL economy. Five-level protection makes mass forgery physically impossible.