The Verdict Is Already In
Twenty-two percent of Americans trust the federal government to do what is right always or most of the time. The figure was 73% when the question was first asked in 1958 and 77% at its peak in 1964 [1]. Asked in 2023 how the political system is working, 4% said extremely or very well; 63% had little or no confidence in its future; 65% said that thinking about politics leaves them exhausted and 55% that it leaves them angry; 28% held an unfavorable view of both parties, the highest share in three decades of polling [2]. Gallup's long series, which averages confidence across the nine institutions it has tracked since 1979, stood at 28% in 2025 against 48% in its first year [3]. The verdict is not American. Across 24 countries surveyed by the Pew Research Center in 2023, a median of 59% were dissatisfied with how their democracy works and a median of 74% said elected officials do not care what people like them think [4].
The institutions that measure democracy itself record the same direction. Freedom House counted 2024 as the nineteenth consecutive year of decline in global freedom, with 60 countries deteriorating and 34 improving [5]. The V-Dem Institute's 2025 report puts 72% of the world's population in autocracies, 45 countries in an autocratizing episode, and the level of democracy enjoyed by the average person back where it stood in 1985, and counts 29 liberal democracies in 2024 [6]. The Economist Intelligence Unit scored the United States 7.85 for 2024, 28th in the world and in the category it names a flawed democracy, while the global average fell to its lowest reading since the index began in 2006 [7].
A system can run on numbers like these for a long time, as the house has written of fiscal arithmetic that cannot be sustained and is sustained anyway. What it cannot do is draw legitimacy from them. The experiment has failed by the measures its own citizens apply, and the citizens have returned a judgment on the machinery. The question this essay takes up is what in the machinery produced the judgment, what other systems have done better and at what cost, and what a reform would have to contain to be more than another cycle of the same quarrel.
Four Flaws Written Into the Design
The flaws are old enough to have theorists. Anthony Downs observed in 1957 that a rational voter has little reason to inform himself, because one vote almost never decides an election and the cost of learning falls on the voter alone, so politics is conducted in slogans [8]. Mancur Olson showed in 1965 that small groups with concentrated interests organize and large groups with diffuse interests do not, so a subsidy worth billions to a few hundred firms is always defended more effectively than the cents it costs each of a hundred million households [9]. Francis Fukuyama gave the third flaw its name: a vetocracy, a system with so many points at which a measure can be stopped that the capacity to decide dissolves while the capacity to block remains [10]. The Congress now sitting has enacted 121 laws since January 2025, the second fewest of any Congress in four decades; its predecessor had enacted 106 by the same point and passed 61% of its entire two-year output in the lame-duck weeks after the voters had spoken [11].
The fourth flaw is money, and it is the one the voters name first. In the 2023 to 2024 cycle, political action committees raised $15.7 billion, congressional candidates $3.8 billion, presidential candidates $2.0 billion and the parties $2.7 billion [12]. Eighty-five percent of Americans say the cost of campaigns makes it hard for good people to run, 80% say big donors have too much influence, 70% say the people in a member's own district have too little, 81% rate Congress poorly at keeping personal finances separate from public work, and 72% want limits on what individuals and organizations may spend [2]. The research behind the complaint is older than the complaint. Martin Gilens and Benjamin Page found that economic elites and organized business groups "have substantial independent impacts on U.S. government policy" while average citizens and mass-based interest groups have little or no independent influence; ordinary citizens get what they want when the affluent want it too, as "more or less coincidental beneficiaries" [13].
Put the four together and the gridlock is not an accident of personnel. A voter rationally uninformed, interests asymmetrically organized, a structure built to block and a candidate pool filtered by its ability to raise money will produce, in any country, a legislature that cannot pass a budget on time, a politics that swings between two poles every cycle, and a class of representatives selected for fundraising rather than for judgment. The quarrel between left and right that fills each cycle is the visible output. The four flaws are the machine.
Two Hundred and Fifty Years
Congress fixed the date itself: "July 4, 2026, the 250th anniversary of the founding of the United States" [15]. The anniversary arrived with the federal debt at $40.3 trillion, a fiscal year that closed with a deficit of $1.992 trillion and a net interest bill of $1.052 trillion, and a budget office projecting that interest will exceed defense [16]. The house has read that ledger before, in The Liquefaction of Forty Trillion and in The Mirage of Fiscal Rectitude, and will not read it again here. What matters for this essay is the shape of the thing, and the shape has precedents.
The white paper Reform Before Rupture set two of them side by side. In the France of the 1780s, Thomas Sargent and François Velde write, "the pressure not to default outran the government's ability to raise taxes", and the taille, the main direct tax, was "a direct tax of medieval origin from which the nobility was exempt": the capacity to pay and the obligation to pay rested with different people [14]. Jack Goldstone, reading the same breakdown over a longer span, named the three systems that failed together, "the inability of its economy, its system of taxation, and its mechanisms of elite recruitment to cope with sustained population growth" [14]. The reader who has followed the house's work will recognize the three. An economy whose demand side has been starved is the subject of The Demand Strike; a tax system that cannot reach the stock of wealth is the subject of a section below; and a mechanism of elite recruitment that selects for money is the fourth flaw above. Peter Turchin's structural-demographic reading extends a sequence of American instability peaks around 1870, 1920 and 1970 to a further peak around 2020, offered as a projection and not a prediction, and names elite overproduction, more aspirants than there are positions to fill, as one of the theory's three principles and its most reliable predictor [14]. Walter Scheidel's survey of how concentration has ended found four levellers, "mass mobilization warfare, transformative revolution, state failure, and lethal pandemics", and no comparable peaceful reform [14].
The house does not forecast, and this essay assigns no date. It states the record: a system of representation that its own citizens rate as described above, inside an economy whose stock of wealth is concentrated to a degree the Federal Reserve's series has not measured before, carrying a debt whose interest now outruns its defense budget, in the two hundred and fiftieth year of its founding document. Systems in that position have evolved or they have ruptured. The record holds no third outcome.
What the Technocracies Got Right, and What It Cost Them
Three states are offered, in the present argument, as evidence that something other than electoral democracy can govern well. The record supports a narrower and more useful claim. Singapore's output per person was $94,897 in 2024 against $86,170 in the United States; it ranked third of 180 countries on Transparency International's 2024 index with a score of 84, behind Denmark and Finland; 77.2% of its resident households live in public housing and 91.2% own their homes [17][18][20]. Its ministers are paid a clean wage with no hidden perks, set by a published benchmark at a 40% discount to a private-sector reference, with every bonus inside the norm rather than on top of it, and the government states the discount as the ethos of public service [19]. The United Arab Emirates scored 68 on the same index, 23rd in the world and five places above the United States at 65, with output per person of $50,274 [17][18]. China, on the World Bank's count, reduced the number of its people living below $1.90 a day by close to 800 million over four decades, close to three-quarters of the world's poverty reduction in that period, from an output per person that stood at $13,293 in 2024 [22][17].
Each of those records carries a price that the same reference works measure. Freedom House scores Singapore 48 of 100 and Partly Free, the Emirates 18 and Not Free, and China 9 and Not Free with a political-rights score below zero; the United States scores 84 and Free [21]. The honest sentence is therefore not that autocracy governs better, but that three states with very different regimes share three things the gridlocked democracies have lost: administrative capacity, a recruitment of the governing class by competence and clean pay rather than by fundraising, and a horizon longer than the next election. The evidence on regimes points the same way. Daron Acemoglu and his co-authors found that democratizations raise output per person by about 20% in the long run [23]. The lesson of Singapore is not that the vote is the problem. The lesson is that the vote was never the whole machine, and that the parts around it, how the governing class is recruited and paid, how far ahead it is made to look and what it is permitted to sell, were left to habit in the democracies and designed on purpose in the city-state.
Every Ism Ends at the Same Table
The present argument holds that every system of the last century has failed, and the record supports a more exact version. Each ism was a rule for distributing what an economy produces, and each rule ended with the stock of capital in a few hands. The command economy ended in 1991 with the dissolution of the state that had run it for seven decades, and the country that did the most to reduce poverty in the decades since did so after loosening that rule. The socialism of the Bolivarian variety ended in the arithmetic of Venezuela, whose real output, on the International Monetary Fund's own series, fell by three quarters between 2013 and 2021 [24]. The Keynesian compact, Keynesian in theory and debt-based in execution as Capital After the Keynesian Model put it, was rolled forward on sovereign deficits normalized as permanent instruments until it met the exhaustion of cheap money. Populism, in the Argentine form The Argentine Mirror examined, ended in a country where INDEC counts 45% of workers outside the formal economy and where a compliant mid-sized firm pays more than its entire commercial profit in taxes and contributions [14]. Capitalism as practiced has not collapsed, and that is the point: it has concentrated. In the second quarter of 2026 compensation of employees was 50.2% of American output, the lowest in a series that begins in 1947, while the wealthiest 1% of households held 32.5% of net worth, the highest in the Federal Reserve's record, and the bottom half held 0.6% of the equities and 51.6% of the consumer credit [25][14].
Five rules produced five endings at the same table. Communism put the stock in the state and the state in a party. Socialism put it in the state and the state in a caudillo. Keynesianism financed the demand side with public debt until the debt and the inflation arrived together. Populism financed it with printed money until the statistics had to be falsified. Capitalism financed it with household credit while the equity went to the top decile. None of them distributed ownership at the point where value is created, the one point at which a distribution does not wait for a budget. That is the gap into which a new paradigm has to fit, and it is a narrow one: a rule that rewards the sources of value, that cannot be withdrawn by a board or a ministry, and that fosters neither the idleness of the handout nor the extraction of the rent.
Why Taxing the Rich Never Did the Distribution
The reflex of the left is to tax the stock and the reflex of the right is to defend it, and both reflexes argue about a flow. The top 1% of American taxpayers earned 22.4% of adjusted gross income in 2022 and paid 40.4% of federal individual income taxes, while the bottom half paid 3% [26]. The same 1% holds a record 32.5% of the wealth [14]. A tax system that collects two fifths of its income tax from the top percentile and watches that percentile's share of the stock reach a record is not failing to tax. Such a system taxes the wrong thing. Income tax reaches income, and the wealth of the top is not income; it is appreciation, unrealized, held through entities. The research on the very top says what happens when a tax system reaches for it. Using leaked offshore records matched to Scandinavian tax data, Annette Alstadsæter, Niels Johannesen and Gabriel Zucman found that the 0.01% richest households evade about 25% of their taxes [27]. The EU Tax Observatory's 2024 report puts the effective tax rate of the world's billionaires at between 0% and 0.5% of their wealth, through holding companies that turn income into untaxed appreciation; the same report records that the automatic exchange of bank information cut offshore evasion by about a factor of three in a decade, and proposes a 2% minimum tax on billionaires that it estimates at $250 billion a year [28]. More than 140 jurisdictions have agreed a 15% minimum tax on large multinationals, with rules in force from 2024 [29].
The honest reading of that record is that loopholes can be narrowed and have been, and that the stock still sits where it sat. The reason is older than the Observatory. Sargent and Velde's taille exempted the nobility by birth; the modern exemption is by structure, by jurisdiction and by the difference between a dollar earned and a dollar appreciated, and every closing of one route has been met by the opening of another or by the departure of the capital to a jurisdiction that did not close it. A tax reaches a flow after the fact and redistributes it through a budget that the fourth flaw above has already captured. The house proposed, in Reform Before Rupture, a distribution that happens before the tax rather than after it: half of a company's net profit to the people who work in it, in equal shares and in addition to salary, written into the fund's legal terms, with a public counterpart the paper calls the Fifteen Compact, a single contribution levied on profit after the workforce's share and rising in bands from 1% to a ceiling of 15%, on the view that an enterprise which has shared half its profit has already done part of the redistributive work a tax system performs [14]. In Argentina, the worked case, the paper's measure put taxes and contributions at 106.3% of a standard medium firm's commercial profit, and under the Compact the personal burden of a registered employee at the median wage would fall from 17.0% of gross pay to 0.3% [14]. Distribution at source is not a tax on the rich but a rule about who owns the profit when it is made, and the one rule none of the five isms tried.
Vote the Policy: What Already Works
The proposal at the center of the present argument is that citizens vote on policies rather than on politicians, through electronic portals, at a fixed cadence, in their own jurisdiction. The proposal is less novel than it sounds, and its precedents are instructive about both its reach and its risks. The Swiss electorate has been called to the ballot box 332 times since 1848 on federal questions, on dates fixed in advance [30]. In Estonia's 2023 parliamentary election 51.1% of participating voters cast their vote over the internet, in a system first used in 2005 [31]. Taiwan's public-policy participation platform, run by its National Development Council, lets any citizen propose a measure; once a proposal gathers 5,000 endorsements the responsible ministry must answer within two months, and may convene a collaborative meeting of petitioners, officials and stakeholders before it does [32]. Ireland's Citizens' Assembly, 99 members chosen by lot, met over five weekends in 2016 and 2017 to consider a constitutional question its parliament had avoided for a generation, and the referendum that followed its recommendation passed with 66.4% [33]. France's Convention Citoyenne pour le Climat, 150 citizens drawn by lot, handed the government 149 proposals in 2020 [34]. The OECD counts 716 representative deliberative processes of this kind in 28 of its member countries between 1979 and 2023, 148 of them in the last three years of that span [35]. Majorities in nearly every country Pew surveyed say direct democracy is a good way to govern [4].
The precedents also mark the design conditions, and Downs's voter is the first of them. A quarterly ballot on which an uninformed electorate chooses between slogans would reproduce the flaw it was meant to cure, which is why the Irish and French cases matter more than the Swiss count: in both, a body chosen by lot deliberated on evidence for months before the public voted, and the public then voted on a proposal that had been examined rather than marketed. The second condition is the floor. A majority of a jurisdiction cannot be allowed to vote the rights of a minority away, and every durable direct democracy sets constitutional questions and individual rights above the quarterly ballot. The third is the jurisdiction itself. A school budget, a zoning rule and a transit line belong to the people who live with them, and a portal that lets a city or a province decide those four times a year, after an assembly has worked the file, takes the gridlock out of the national legislature by taking the decisions out of it. What the portal cannot do is replace judgment with a poll. What it can do is end the practice of electing a person every four years on a bundle of positions and then discovering which of them was for sale.
Vet the Politician, Separate the Money
The second proposal is that politics become a vetted career: entry by education and examined competence, continuation by a verified record of integrity, pay set in public and clean, and no selection by the size of the war chest. The precedent for that, too, already exists. Singapore sets its ministers' salaries by a published benchmark with a 40% discount to the private-sector reference, states the discount as the ethos of public service, pays a clean wage with no hidden perks and counts every bonus inside the norm, which the 2012 white paper fixed at S$1.1 million a year for an entry-level minister while cutting the Prime Minister's salary by more than a third [19]. One may dispute the number and keep the method: a pay rule published in advance, benchmarked, discounted, audited, with nothing hidden. That method is the one the house wrote into Tenet 2 of The SAVI Capital Model for the chief executives of the companies it finances, a ceiling on pay as a multiple of the lowest wage, and into Tenet 3, stewardship that an outsider verifies every year. A legislature that applied the same two rules to itself, a published ratio between a representative's total compensation and the median wage of the district, and an annual audit of assets and interests by a body the representative does not appoint, would have answered the 81% who rate Congress poorly at keeping its finances separate from its work [2].
The vetting of integrity is harder than the vetting of competence, and the honest statement is that no examination certifies righteousness. What can be certified is the record: a published history of assets, of votes cast against disclosed interests, of the ratio above. The design principle the house applies to a fund applies to a legislature: what is expressed can be withdrawn and what is encoded cannot, and a governance that lives in speeches is theater. Separating money from politics follows from the same principle, and the remedy has been tried at municipal scale. Seattle's voters created a Democracy Voucher Program by initiative in November 2015, under which every eligible resident receives four $25 vouchers to give to the candidates of their choice, and in 2025 approved a property tax of $4.5 million a year to fund it for a further ten years [36]. A voucher does not ban a donor. It dilutes one, by handing every citizen the same stake in the campaign that the donor buys. Fifty-eight percent of Americans believe laws can effectively reduce the role of money in politics [2]. The research says they are right about the diagnosis, and the Seattle vote says a city can act on it without waiting for a national legislature that the money elects.
The Middle Ground Is a Mechanism, Not a Mood
Each of the reforms above addresses one of Goldstone's three systems: the ballot and the vetted career address the recruitment of elites, and distribution at source addresses taxation. The third system is the economy, and here the house's position is the one the white paper states. The quarrel between the haves and the have-nots is settled neither by a mood of moderation nor by a coalition that splits the difference every cycle, but by a rule that pays both from the same profit at the moment the profit is made. The firm commits to write into the legal terms of every fund it raises that half of a financed company's net profit, after the audited result, belongs to the people who work there, in equal shares and in addition to salary, and half to the capital that financed them; that executive pay is bound to a ratio; that stewardship is verified; and that returns above five times investors' contributions pass to an endowment [14]. The design rewards the sources of value and nothing else. The pool is paid out of audited profit, so a company that creates none distributes none; it is paid in equal shares, so the apex cannot capture it; it is paid in addition to salary, so it is not a wage disguised; and once written into the fund document, neither a board nor a ministry can withdraw it in a bad quarter. It fosters no idleness, because the enterprise earns it, and no handout, because the state is not the payer. The cost to investors is stated in Appendix C of the paper, a lower multiple on the same outcome, and the paper claims no outperformance in exchange, because no study yet tests sharing at half of profit [14].
The glossary names the shapes this produces. The present distribution is the K-Shaped Society; the mechanism that produced it is the Asset Loop; the alternative is Structural Inclusion, ownership embedded in the architecture of the enterprise; its result is the Symmetric Middle; and the debt owed by a system that extracted for a generation is the Moral Debt the model is built to repay. A legislature that wanted the quarrel to stop would act in this order. It would offer the Fifteen Compact as an opt-in regime, so that enterprises which share half their profit are taxed once and lightly and the budget stops compensating for what the enterprise withheld. It would put the jurisdiction's choices to its citizens on a fixed cadence, after a body drawn by lot has worked the file. It would publish the ratio and audit the record of the people who serve it, and pay them a clean wage. And it would hand every citizen a voucher the size of a donor's cheque. None of these is a measure of the left or of the right. Each takes a decision out of the hands that have been selling it.
Churchill told the Commons in 1947 that democracy had been called "the worst form of Government except all those other forms that have been tried from time to time" [37]. The sentence has carried the system for seventy-nine years, and the record above says it has carried about as far as it will. The citizens have judged the machinery; the index-makers have measured the decline; the ledger has posted the interest line; the demographers have counted the cradles. Reform Before Rupture was written as an argument. The two hundred and fiftieth year has made it a choice.