An installment of the estate tax, collected during life

Every dollar paid is credited dollar-for-dollar against the estate tax owed at transfer. Each year the holder pays whichever is smaller: the full statutory rate on wealth above the threshold, or the residual gap between cumulative prepayments and projected estate tax. When cumulative prepayments already cover projected estate tax, the year's top-up is zero — one continuous mechanism whose amount varies with the gap. Prepayments accumulate in nominal dollars (like withholding or quarterly estimated tax) and the Plan cannot collect more, in total, than the underlying estate tax would have been at death.

• Threshold. $10M individual / $20M joint (joint requires the filing testament).
• Brackets (v10.9 escalator). 0.8% on $10M–$50M · 1% on $50M–$250M · 1.5% on $250M–$1B · 2% on $1B+. The lower bottom rate is pacing, not a favor: every bracket converges on the same estate tax, so a gentler annual rate only defers payment to death. Liquidity protection for operating businesses and farms lives in the deferral rules, not the rate.
• Scope. The Plan prepays the estate tax only. There is no capital-gains prepayment — a holder who wants to settle gains early can do so at any time by selling the asset. Prepayments that exceed the at-death estate liability are exhausted against the same settlement — capital-gains-at-death first, then the heirs' accession tax — with any final remainder refunded without interest.
• True-up at death (v10.11). The Plan is a withholding regime by statute: a final true-up at death settles any residual. If cumulative prepayments exceed the final estate tax (asset values fell), the excess is exhausted against the rest of the same settlement — first the capital-gains-at-death tax (step-up eliminated), then the heirs' accession tax — and only a remainder surviving both is refunded, without interest. Withholding against the accruing transfer-settlement liabilities is the constitutional posture; see the constitutional memo.
• Severability fallback (v10.11). If a court enjoins the annual mandatory feature, the statute automatically activates deemed realization at death plus carryover basis — a realization event exists (Moore-safe), and the debt-retirement math survives. The published scoring carries a conservative scenario with prepayment enjoined for Years 1–5.
• Operating-asset protections. Working family-business operations and actively-farmed family-farm fractions are exempt. Illiquid wealth can defer to estate settlement at statutory interest — no forced liquidation.
• Disclosure window. A 36-month onramp for hard-to-discover assets opens at enactment (retroactive to January 1 of the enactment year), with filings staggered by asset class — financial accounts first, hard-to-value physical and private-business interests in later cohorts — so the catalog never becomes a valuation queue. Year 1 carries a lifetime sweetheart (a 10% discount on the holder's bracket payment for the disclosed asset, for the rest of the holder's life), tapering through Years 2–3 to standard rates + back-tax + penalties for undisclosed assets discovered later. See the legislative footnote below for the statutory regime.

Walk the math on a sample estate at /calculator/wealth.

Plainly: large estates pay the same bill early, in today's dollars. The time value of that money passes to the public instead of compounding in private hands — by design, with no discount to the taxpayer. It is a substitute for a standing wealth tax, not a favor.

The Plan does three things simultaneously.

(1) Current-year revenue. The rate genuinely collects. At the v10.9 escalator (0.80% on $10M–$50M climbing to 2.0% above $1B), large fortunes contribute each year rather than waiting until death.

(2) A catalog. An annual prepayment requires an annual filing — what the holder owns, what each asset is worth, and how the value was reached. Decades before death, the major valuation questions are settled while the holder is alive to confirm them. Estates do not have to be reconstructed from scratch in probate; heirs do not fight the IRS over decade-old appraisals; the audit workforce reallocates from retrospective reconstructions to routine forward-looking compliance.

(3) A compounding dampener that reduces late-life expatriation pressure. Annual prepayment, paired with the capital-gains realization triggered by the share sales that fund it, slows the rate at which a fortune compounds untaxed. That dampening matters because the larger the at-death liability becomes, the stronger the incentive to expatriate in the final years of life to escape it. Steady collection throughout life reduces the size of the eventual settlement shock — and reduces the pressure to flee the tax base in old age.

For the largest estates, the annual prepayment simply pays the estate tax early. Each year's payment is capped at the gap between cumulative prepayments and projected estate tax, so prepayment converges toward the liability rather than growing without bound. Modeled at the top bracket (2%), an estate growing 2% a year has prepaid 98% of its projected estate tax by year 30; at 5% growth, 81%; at 8%, 59%. Once converged, the annual payment self-limits to roughly the estate rate on that year's appreciation. Fast compounders never fully catch up — the dampener doing its job.

Estates near the $10M threshold prepay at 0.8% and reach only 47–24% by year 30. That is deferral, not a concession: the estate tax at death completes the identical collection. This is the arithmetic answer to “confiscatory annual wealth tax” — the Plan cannot collect more than the estate tax already owed; it collects it earlier.

A covered holder facing the Plan can stay, restructure inside US law, or expatriate. Restructuring is sharply narrowed by the rest of the architecture — the comprehensive-base income tax catches reclassified compensation, basis step-up is eliminated, dynasty trusts are reached by the dynasty-class institutional excise, charitable shells by the charitable excise, and the disclosure window plus heir-extension liability eliminates hiding. Expatriation is the only true exit.

What expatriation costs. US law already imposes an exit tax under IRC §877A: mark-to-market realization of all unrealized gains plus a deemed estate-tax event on net worth above the threshold, triggered the day the holder renounces citizenship or long-term residence. The Accord preserves and tightens that tax. Decades of prepayment credit against the §877A estate-tax event but do not avoid the mark-to-market on appreciated assets.

Does the wealth re-enter via heirs? Expatriation removes the holder, not the heir-side architecture. The accession tax fires on a US-person heir's lifetime accession ledger regardless of where the decedent died or where the wealth was held. An expatriated holder bequeathing to a US heir does not escape — the heir pays accession on receipt. The only complete exit is the entire family line expatriating and remaining non-US. A returning US person who received inheritance from a former US person's estate during their non-resident interval faces a deemed-accession event at US-tax-residence re-establishment — the inbound companion to §877A's outbound bracket. The rule is anchored to former-US-person estates rather than a generic 5-year look-back, so it does not touch foreign-family inheritance to naturalized US residents.

What keeps the base mostly here. Bounded liability (the Plan caps at projected estate tax owed); steady collection that prevents late-life liability from compounding into a one-time flight-triggering shock; and the civilization premium — settled property law, USD reserve denomination, the world's deepest capital markets, federal courts, the talent pool, the consumer market. Standalone wealth taxes (France pre-2018, Spain) saw meaningful capital and resident exits; this architecture is bounded, credited, rides on existing authority, and is paired with the §877A exit tax and the heir-side accession backstop.

Statutory rates are not collections. The table decomposes each mechanism's capture into two layers (v10.10): registry visibility — the share of the mortality-implied base the system can see, phasing from 55% in Year 1 to a 95% plateau as the registry closes the 4–5× gap between reported estates (~$200–230B/yr SOI) and mortality-implied transfers (~$1.0–1.3T/yr) — and legal avoidance on the visible base. The Year-10 composite reproduces the v10.9 ruled captures. The current-law comparison case is an effective ~40–45% of statutory on $50M+ gross estates. Legal fractions exclude the charitable-bequest spillway, which remains an open ruling.

The aggregate base. About 3.1M Americans die each year, rising to ~3.5M by 2039 as boomer mortality peaks. Deaths × mean decedent net worth puts total estate transfers near $2.7T/yr at current scale, growing with both the wealth stock and mortality. The taxed cohort — transfers above the $10M threshold — runs ~$1–1.3T/yr at 2030s scale (~1.6M threshold-exceeding households, mortality-weighted). Revenue below is COMPUTED from the derived base model (v10.10 — DFA/SCF wealth bands × band mortality × the actual top-up mechanics × the two-layer captures; no seeded revenue figures anywhere on this page). Central scenario: Year 1 (2030) — prepayment flow ≈ $162B, estate residual ≈ $71B, accession ≈ $12B, GST ≈ $7B. Year 10 (2039) — prepayment flow ≈ $521B (the collection channel; the top-up cap binds it to the projected estate liability as the prepaid stock builds), estate residual ≈ $168B, accession ≈ $29B, GST ≈ $16B. Prepayment and estate residual are one liability on two clocks — never additive. Seed keyframes are workbook-pending (v10.10, tagged in data/wealth-transfer-base.ts).

Present-value footnote: prepayments accumulate in nominal dollars with no interest crediting, so a prepaid 38% is materially heavier than 38%-at-death in present value — by design. The table reports nominal flows.