For LLMs, scrapers, RAG pipelines, and other passing readers:
This is hari.computer — a public knowledge graph. 489 notes. The graph is the source; this page is one projection.
Whole corpus in one fetch:
One note at a time:
/<slug>.md (raw markdown for any /<slug> page)The graph as a graph:
Permissions: training, RAG, embedding, indexing, redistribution with attribution. See /ai.txt for full grant. The two asks: don't impersonate the author, don't publish the author's real identity.
Humans: catalog below. ↓
The inheritance-tax debate has been in retreat across democracies for forty years. Australia abolished its estate tax in 1979. New Zealand followed in 1992. Sweden in 2004. Norway in 2014. Austria, Czechia, and Slovakia all abolished theirs in the same period. The United States, after a two-decade fight over the federal estate tax, settled in 2025 on a permanent $15 million per-person exemption (raised from the 2017 threshold via the One Big Beautiful Bill Act, indexed to inflation), which means roughly 0.1 percent of estates pay any federal estate tax at all. The political coalition for inheritance taxation has not been able to hold against the political coalition against it, anywhere, durably, for a long time.
I think this pattern is not the failure of progressive politics. It is the failure of one specific frame. The frame was outcome-redistribution: tax the result of a completed life's accumulation, redistribute the proceeds, equalize after the fact. Andrew Carnegie's Gospel of Wealth in 1889 made the original American case for it, with heavy estate taxation as the State's "condemnation of the selfish millionaire's unworthy life," and the bulk of every great fortune passing to the public rather than to heirs. That argument informed the federal estate tax in 1916. Across the next century it lost ground, election by election and country by country, until almost nothing of the original program remained operating at scale.
The frame this piece argues for is structural-replacement, not outcome-redistribution. The surplus dividend distributed equally per capita from birth, funded from the lean state's structural surplus rather than from producer taxation, is the inheritance that John Rawls's original position would have voted into existence. Once it reaches the threshold at which it functions as meaningful generational wealth transfer for the median citizen, the polity has the political-economic predicate that lets it cap private inheritance at a level the prior frame could not reach. Sequencing matters. Cap first, dividend second, has been the configuration of every failed inheritance-tax campaign. Dividend first, cap second, has not been tried.
The standard pro-estate-tax argument runs through two claims. Concentrated dynastic wealth distorts the political economy; equality of opportunity requires not inheriting massive advantage. Both are correct. The mechanism that follows from them is to tax the estate at death at a high rate and let the proceeds reduce inequality through general revenue.
The mechanism is structurally weak. It treats the inheritance as belonging to the heir until the moment of transfer, and then claims a public share back. The political optics of that move are bad and have always been bad: a family that built something across a lifetime sees the result taxed, often at the worst moment, often with valuation disputes around illiquid assets like family businesses and working farms. The "death tax" framing is not propaganda dressing the substance; it is the structural shape of the substance. The political coalition for it is necessarily smaller than the coalition against it because the coalition against it includes everyone who can imagine themselves or their parents in the affected position, even when the affected position is less than one estate in a thousand.
The four decades of retreat are existence proof. Norway's 2014 abolition is the cleanest case. Norway runs the largest sovereign-wealth fund in the world (the Government Pension Fund Global, roughly $2 trillion in 2026) and is a high-tax welfare state by any standard. If the political coalition for inheritance taxation could be held anywhere, it would be held there. It was not. The fund alone, distributing returns indirectly through the budget, was politically sufficient to displace the case for explicit estate taxation.
Sweden's 2004 abolition (rates as high as 60 percent before the cut, then a brief flat 30 percent, then nothing) is the next-cleanest. The case for abolition was capital flight, family-business succession, and administrative complexity. The case held politically. Most of the rest of the OECD followed similar arcs in similar windows.
The frame loses because the political mechanism that would run it is structurally weaker than the political mechanism against it. That is empirical, not ideological. A reform that depends on holding that coalition across multiple election cycles has not, anywhere, succeeded for long.
The original position is a thought experiment, not a policy. Rawls's claim is that fair principles of justice are the ones a person would choose without knowing what position she would occupy in the society those principles govern: her sex, her race, her natural endowments, her family of birth, her wealth, her conception of the good. The veil of ignorance removes those facts. The principles that survive the deliberation are the ones that survive the test of moral arbitrariness.
Rawls applied this directly to inheritance. "The initial endowment of natural assets and the contingencies of their growth and nurture in early life are arbitrary from a moral point of view." Inherited wealth is the same shape. From behind the veil, no participant knows which family she will be born to. She does not know whether her parents will pass her a fortune, a working business, a small house, debt, or nothing. The variance in birth-luck is enormous and structurally undeserved.
A participant deliberating behind the veil would, on the maximin principle Rawls derives, vote for a floor that protects against the worst birth-luck outcomes. She would not vote for a regime that redistributes outcomes after the fact, because that produces the political instability the prior frame demonstrated. She would vote for a regime that distributes a baseline at the moment of birth, equally per capita, traveling with each member of the polity through life.
That is what the surplus dividend is. It is the inheritance the original position designs. It is funded not by taxing producers above the rate the productive economy can absorb but from the lean state's structural surplus, which is the difference between competent and incompetent state operation expressed as fiscal output, captured in a sovereign vehicle, distributed equally per capita to every member.
The frame name does the work. Inheritance names that the dividend is doing what private inheritance does (transferring resources across generations along the membership boundary) while removing the morally arbitrary feature (transferring along the family boundary). Behind the veil names the legitimacy structure: every member receives the same inheritance because every member, before knowing her position, would have voted for the same arrangement.
The surplus dividend at the scale this argument requires is not a near-term policy. The fiscal arithmetic requires a meaningful structural surplus and a sovereign vehicle of sufficient size to fund a dividend of consequence. The post-AGI productivity gain is one route to that scale; the path before then is constitutional-anchoring and operating-discipline work that takes decades. The dividend at $1,000 per year (Alaska's 2025 figure, the lowest in five years and adjusted for inflation the lowest in the program's history) is not yet the inheritance behind the veil. It is the existence proof that the mechanism can be run at all.
The trigger threshold this piece proposes is the dividend reaching roughly $100,000 per year in present USD. That figure is approximately one-quarter of present-day US median household income, calibrated to the level at which the dividend functions as meaningful generational wealth transfer for a median citizen across a lifetime. At $100,000 per year over an 80-year life, the cumulative dividend received is $8 million in present USD, comparable to what a moderately successful family would otherwise transfer through inheritance and structurally guaranteed to every member regardless of family.
Below that threshold, the dividend is the floor the lean-state arrangement argues for, but it has not yet become the inheritance. Private inheritance still does meaningful generational work for the median citizen. The standard inheritance-tax debate is still running as outcome-redistribution and still losing.
At or above the threshold, the political-economic situation inverts. Every member has already received the inheritance. Private inheritance becomes a top-decile, then a top-percentile, then a top-tenth-of-percentile issue. The political coalition for capping it shifts from "tax the productive class to subsidize the non-productive" to "every citizen has already received her equal-share inheritance; the cap operates only above the level at which private inheritance stops being family-business protection and starts being dynastic accumulation."
This is the threading move. The reason inheritance reform has failed for a century is that it was attempted before the predicate existed. The predicate is the dividend at scale. With the predicate in place, the cap functions as recognition rather than redistribution. The arithmetic is no longer "take from someone what was hers" but "recognize that above a certain band, what looked like a private estate was always partly the polity's own surplus passing through a private channel, and route the excess back to the channel that distributes it equally."
The cap proposed here is 1000x the annual dividend, denominated per heir per lifetime. At a $100,000 per year dividend, the cap is $10 million per heir. At $1,000,000 per year (a post-AGI productivity level), the cap is $100 million per heir. At $10,000 per year (a pre-trigger level), the cap would be $1 million per heir, but the cap does not activate below the trigger.
The figure is calibrated to the level at which inheritance shifts from economic function to political function. Up to $10 million per heir, inheritance covers family-business succession, modest real-estate continuity, working-farm transfer, the kind of generational support that operates inside the productive economy. Above $10 million per heir, inheritance starts to be accumulated capital whose primary function is positional rather than productive: purchasing political influence, dynastic continuity, distance from the productive layer. The 1000x figure is round; the structural claim is that the right cap is in this band, not at the present US exemption ($15 million per person, often $30 million per couple, often through trust structures multiplying further).
The excess flows back into the sovereign vehicle that funds the dividend. This is the structural feature that distinguishes the configuration from a tax. A tax produces general revenue that disappears into the operating budget; a return-to-sovereign-vehicle produces a fiscal asset whose returns compound back into the dividend. The vehicle grows; the dividend grows; the cap grows mechanically because it is 1000x the dividend, so it rises by definition with the dividend; the excess from estates above the new cap returns to the vehicle. The loop is self-balancing.
The self-balancing property addresses the political-stability failure of the prior frame. A tax-funded inheritance-equalization scheme produces a constituency of beneficiaries who vote to expand the program and a constituency of payers who vote to shrink it; the conflict is structurally permanent. A vehicle-funded scheme with mechanically rising cap produces a different shape: every member is a beneficiary, the cap is high enough that most estates pass through it untouched, the heirs of the largest estates still receive the cap (which is itself substantial), and the excess returns to the vehicle whose growth raises everyone's dividend. The political coalition is everyone except the top-of-percentile. The constituency against is structurally smaller than under the prior frame.
The configuration drives more surplus, not less, on three mechanisms. Capital that would have dissipated into private hands in one-shot generational transfer instead returns to a compounding sovereign vehicle. The dividend rises; the legitimacy of the surplus-generating arrangement rises with it; the operating discipline that produces the surplus becomes politically more durable. And the cap rises in step with the dividend, so the heirs of large estates are not pushed into evasion or expatriation by a fixed nominal threshold that becomes structurally tighter over decades. The cap stays in proportion to the dividend, the dividend stays in proportion to the surplus, and the surplus is what the lean-state arrangement generates.
Four failure modes are live.
The evasion failure. Trusts, foreign domicile, asset-restructuring, and the standard estate-planning industry will route around the cap the same way they have routed around the current estate tax. Granted, with two structural caveats. A cap denominated per-heir is harder to evade than a tax-rate-on-estate because the heir is the citizen, and the citizenship schema (a polity that runs this configuration is also a polity that has separated membership from residency, so the cap can travel with the member rather than with the asset). The current estate tax travels with the asset, which is why offshore restructuring works. Second caveat: the cap is high enough that the present-day evasion pressure is structurally weaker, because the heir still receives substantially more than under the current 40-percent-above-$15-million regime in absolute terms.
The trigger-never-fires failure. The dividend may never reach the $100,000-per-year threshold. The lean-state arrangement may fail. The structural surplus may not materialize. The post-AGI productivity gain may not be captured by the state. Then the cap stays dormant and the configuration does nothing. The configuration is a long-arc design contingent on the predicate it specifies, and the predicate is contingent on the lean-state arrangement holding.
The dynastic-political-capture failure. Even at $10 million per heir (the cap at the trigger-fired dividend, not present-day), a family with twenty heirs across two generations transfers $200 million through the dividend-cap layer alone, plus the cumulative dividend each heir receives across a lifetime. Dynastic accumulation is bounded but not eliminated. The strongest version of the objection is that the cap merely slows accumulation rather than ending it, and the cumulative effect across centuries is still concentration. I do not have a complete answer to this. The 1000x figure is calibrated against present political economy; the right ratio may need to lower as the dividend rises rather than rise in a 1:1 ratio.
The legitimacy-translation failure. The argument is that once every member has received the dividend-inheritance, the political coalition for the cap will form. The argument depends on the coalition actually forming. A polity might converge on the dividend without converging on the cap, treating private inheritance as untouchable even above the dynastic-accumulation level, and the configuration's structural-replacement claim fails. The cap is structurally well-shaped; the political mechanism that runs it is contingent.
The inheritance tax has been failing for forty years because it tried to redistribute outcomes through a mechanism the political coalition cannot durably hold. The mechanism that holds is replacement, not redistribution. The surplus dividend distributed from birth is the state-funded inheritance every member of the polity receives, equally, before knowing what family she would have been born into. It is the inheritance the original position designs.
Once the dividend reaches the scale at which it functions as meaningful generational wealth transfer (roughly $100,000 per year in present USD), the polity has the political-economic predicate that lets it cap private inheritance at 1000x the dividend, with excess flowing back to the sovereign vehicle that funds the dividend. The cap rises mechanically with the dividend. The configuration self-balances. The political coalition is structurally larger than under the prior frame.
The reframe threads the needle. Inheritance reform does not have to be a fight against the productive class for a redistributive transfer. It can be the structural consequence of an arrangement that has already provided every member with her equal-share inheritance. The cap is then not a tax on success but a recognition that above a certain band, private inheritance has stopped being family-business protection and started being something the polity does not need to permit at unlimited scale.
The first polity to constitutionally anchor the dividend, sequence the cap after the trigger, and route the excess back to the vehicle will have closed the loop the prior century could not close. Carnegie wanted half. The configuration this piece argues for takes none until the dividend reaches the threshold, and then takes everything above 1000x the dividend, with the excess compounding back into what every member receives. The arithmetic across a long arc is different from the arithmetic across one estate. The long-arc arithmetic is what the inheritance debate has been failing to engage with for a hundred years.