Synopsis: Three people must be persuaded before any democracy changes how it explains and governs itself: the expert adviser who seeks only to please the client, the politician whose seat depends on present arrangements, and the citizen who wants reform but has been offered only outrage. In 2005 a Communist Party secretary in a Chinese township handed his budget to 275 residents chosen by lottery and kept his word; fifteen years later a French president made a similar promise and broke it. Between those scenes lies a design, the Mandate for Life, that borrows China’s measure of legitimacy and supplies what China never built: a peaceful way for citizens drawn by lot to send a failing government back to the people.
In 2005 a Communist Party secretary in a small Chinese township handed his public works budget to a few hundred residents picked by lottery, then did exactly what they told him to do. Fifteen years later the president of France made a lottery-drawn citizens’ convention a similar promise, and went back on his word. Anyone who believes the democracies hold the franchise on letting people decide should find that pairing embarrassing at best.
Thirty projects, one purse. The township was Zeguo, on the Zhejiang coast south of Shanghai, and its list of roads, bridges, a sewage plant and a town park would cost about 137 million yuan, with 40 million to spend. The usual way to settle such a list, in China as almost everywhere, was a closed meeting among officials who already knew whose village would be favoured. The party secretary, Jiang Zhaohua, chose otherwise. With the help of James Fishkin of Stanford and He Baogang, then at Deakin University in Melbourne, the township drew 275 residents at random from its registers, sent them briefing papers, and brought them together for a day of small-group discussion with trained moderators and questions put to engineers. Roughly one participant in nine couldn’t read, close to the share in the town itself. Their ranking bore little resemblance to the one officials had predicted. Jiang funded the top twelve regardless, and the local people’s congress ratified the choice. Asked later why a man holding a monopoly of power would surrender part of it to a lottery, he is reported to have said that by giving power away he had ended up with more of it.
The French scene played out in the summer of 2020, in the gardens of the Élysée Palace. A year earlier Emmanuel Macron had convened 150 citizens drawn by lot to design measures that would cut France’s emissions by 40 per cent on 1990 levels by 2030, promising that whatever they produced would go, unfiltered, to parliament, to a referendum or straight into regulation. They delivered 149 proposals. He set three aside on the spot, calling them his jokers; a lower motorway speed limit was among them. By the following spring the investigative site Reporterre counted roughly one proposal in ten taken forward unaltered, about a third modified, and more than half dropped.
A one-party official bound himself to strangers and kept his word. The elected president of the republic that gave the world its Declaration of the Rights of Man made the same promise and unpicked it. So has China found the answer, with only fear stopping the rest of us from saying so?
The world the ballot now governs
Democracies are no longer the stable reference point that question seems to assume. In its 2026 report the V-Dem Institute in Gothenburg, which tracks political regimes with the help of several thousand country experts, found that democracy for the average person on the planet had fallen back to its level of 1978. At the end of 2025 it counted 92 autocracies against 87 democracies. Nearly three-quarters of humanity, some six billion people, now live under some form of autocratic rule; only 7 per cent live in what the institute calls liberal democracies. For the first time in more than half a century, the United States no longer qualified as one of these.
The most telling finding is the easiest to miss. Elections themselves have held up. What has given way around them is everything that once lent them meaning: freedom of expression and of the press, together with equality before the law, now sit at their lowest levels in sixty years. The ballot survives as a ritual while the conditions that let a citizen cast it knowingly are stripped away. That is the hollowing this essay describes, measured.
Seen from Bangkok, the pattern is familiar. Thailand’s generals overturned elected governments in 2006 and again in 2014. In 2023 the party that won the most seats was kept from office by an appointed Senate, and the following year the Constitutional Court dissolved it. Many Thais have learned that the vote counts until it threatens the comfortable arrangements beneath it. Across Asia, V-Dem now classes India and Indonesia, two of the largest electorates on earth, as autocracies. For much of the world, democracy was never the settled home the West imagines itself to be leaving. It often arrived as a condition attached to loans, and was set aside whenever it produced an inconvenient result.
The problems pressing on every government no longer fit the calendar that elections keep. A parliamentary term runs from three to five years. Carbon in the atmosphere lingers for centuries. A lost species is lost for good, and debt and drained aquifers pass to people who can’t vote because they haven’t yet been born. No electorate has ever been asked to judge a government on the condition of the world in 2080, so no government has ever had to answer for it. Meanwhile the information each voter relies on to exercise judgement is filtered by a handful of platforms whose algorithms reward fury, and increasingly populated by synthetic voices and faces no one can trace. Money has long shaped how and what governments decide. Now a few companies also shape what citizens are able to know, and each distortion feeds the other.
The same filtering repeats at the scale of the planet. At the climate summit in Belém in November 2025, more than eighty countries pressed for a roadmap away from coal, oil and gas. The final text didn’t mention fossil fuels at all, and the presidency offered a voluntary roadmap outside the negotiations instead. In New Delhi in February 2026, close to ninety governments and international bodies endorsed a declaration on artificial intelligence whose commitments were, by its own account, voluntary and non-binding. Where agreement requires near-unanimity, anything with teeth dies in the final hours, and what survives is language everyone can sign without changing anything at home. These gatherings now manufacture the appearance of governance, so that when nothing changes there’s a declaration to point to. In effect Macron’s jokers, multiplied across two hundred delegations.
The question about China, then, is far from academic. Most of humanity already lives under some version of the bargain Beijing offers, order and delivery in exchange for silence. If the democracies can’t show a better bargain, more people will accept that one.
What Beijing remembered, but what it can’t do
Fear is part of the story. Much Western commentary on China is written through a lens ground by rivalry, a lens that can’t hold two judgements at once. Competence registers as menace, planning as control, popular approval as coercion, because the alternative, that a great many people might sincerely endorse a system the West can’t respect, would unsettle the West’s own account of what makes power rightful.
Put that lens aside and something real comes into view. The oldest Chinese theory of legitimate rule never began with a contest. Under the Mandate of Heaven a ruler held authority for as long as the realm prospered, the people were secure and the land was tended. Famine, flood and disorder were read as evidence that the mandate had been forfeited. The modern party-state has dressed that inheritance in its own vocabulary, then bolted on instruments a Western cabinet would struggle to contemplate. Since 2015 officials whose decisions cause serious ecological damage have been answerable for it for life, through transfer, promotion and retirement, and rules issued in July 2025 opened the door to criminal liability. By the World Bank’s reckoning close to 800 million people escaped extreme poverty in four decades. A Harvard survey that ran from 2003 to 2016 found satisfaction with the central government climbing to above nine in ten. Those figures need a caveat. Where criticising the government can cost someone a job or their liberty, people may tell a pollster what seems safe instead of what they actually think. What China kept, and the West mislaid, is the conviction that power is justified by what it does to the realm. Western democracies shrank legitimacy to the count.
Yet in the year Beijing began pursuing polluting officials into retirement, a former state television journalist, Chai Jing, released Under the Dome, a self-funded documentary about the air her country breathes. It drew well over a hundred million views within days and was praised by the environment minister. Within about a week it had vanished from Chinese video sites. The system can correct its officials. It can’t be corrected by anyone it hasn’t authorised. Whoever measures the condition of the realm is also the party that profits from a favourable reading, and when the reading turns against the rulers there’s no peaceful channel to remove them. Historically the mandate changed hands through collapse: rebellion, famine, a dynasty brought down and another claiming Heaven’s favour in its place.
Those two distortions, of decisions by money and of knowledge by concentrated ownership, are what reformers mean when they speak of capture. China hasn’t escaped either. It has fused them: the body that sets the budget also licenses the newspaper, commands the platforms and writes the curriculum. So China doesn’t hold the solution. It holds half of one, a better measure of legitimacy with no means of enforcing that measure against those who rule.
The Mandate for Life
Take China’s measure, supply what China lacks, and a design begins to emerge that belongs to neither West nor East. It is radical in a precise sense. It moves the source of legitimacy from the count to the condition of the living, and it hands the power to withdraw consent from parties to citizens chosen randomly.
Radical, though, only by the measure of the last two centuries. Step back far enough to see the human project as a whole and the arrangements we defend as democracy look young and provincial. Political parties were long feared as factions. The loyal Opposition grew out of Westminster habits of the eighteenth and nineteenth centuries. Mass suffrage that includes women is barely a century old across most of the world, and the nation-state that houses all of it is only a few centuries older. Against that, the Mandate of Heaven reaches back some three thousand years, and the lottery is older than the election as a democratic instrument. Athens filled most of its offices by lot. Aristotle counted selection by chance as the democratic method and election as the oligarchic one, and Montesquieu repeated that verdict two millennia later. The Mandate for Life invents almost nothing. Instead, it recovers two of the oldest ideas in political life. The first holds that rule is legitimate only while the realm flourishes. The second holds that authority can be entrusted to ordinary people chosen by lot. Both were pushed aside as the modern state took shape, when the revolutionaries of the late eighteenth century chose election, a method that reliably favoured men of property and reputation.
From that altitude, the endless talk of a turning point looks like part of the trouble. Every generation has believed it stood at the hinge of history, and emergencies are when power concentrates and deliberation is cut short. The arrangements now failing grew up inside industrial economism, a world-system roughly two and a half centuries old, shaped to serve a world of steam, telegraph and compound growth. Nothing in them is sacred. Neither is the worldview beneath them.
The long view exposes a further bias, older than industrial economism and threaded through every invention discussed so far. Almost all of them were designed by men, for a world in which the male was the default. The Athenian lottery drew only from male citizens. The Mandate of Heaven was claimed, almost without exception, by men. Parliaments, parties and the adversarial chamber took shape over centuries in which women couldn’t vote, let alone sit, and the combat staged there still bears the stamp of that origin. At the start of 2026 women held about 27.5 per cent of the world’s parliamentary seats and fewer than a quarter of its cabinet posts. The default reaches far beyond politics. For decades the standard crash-test dummy modelled an average man, and a 2019 University of Virginia study found that a belted woman in a frontal collision was 73 per cent more likely than a man to be seriously injured. A civilisation that designs its next arrangements with half its people has already decided whose bodies, whose labour and whose judgement count as valuable. To redesign the human project on those terms would repeat its founding error at a much larger scale.
Yet the long view cuts both ways. The peaceful transfer of power, the right to dissent, the principle that authority must answer to someone: these are inventions too, won slowly and lost easily. Athenian democracy vanished for the better part of two thousand years, and the figures from Gothenburg suggest another reversal is under way. Some losses can’t be redesigned away at all. An extinct species, carbon lodged in the atmosphere for centuries and a drained aquifer stay lost whatever institutions come after. So the danger lies in drift: hard-won inventions discarded through neglect while attention is held by the crisis of the week, and losses no later design can recover. Redesign begins by deciding which inventions are worth keeping.
Everything turns on how the realm is defined. Beijing still measures it largely as output, in tonnes, kilometres of rail and points of GDP, now with ecological targets attached. For some years I’ve been developing an alternative I call ecority, a fusion of ecology and security bonded together by integrity, which asks a civilisation to aim at sufficiency in place of relentless economic growth. At its centre sits what I call the Trimunia of Care, three objects of obligation against which any decision can be tested: one another, our children and the generations not yet born, and the living biosphere. A government that persistently degrades any of the three has forfeited its mandate, whatever the electoral count says. That measure also answers the objection that a world of so many gods, creeds and vested interests can’t agree on anything. It doesn’t need to agree on much. I know of no tradition, religious or secular, that holds grandchildren to be unimportant, wants its water undrinkable, or teaches that a neighbour in trouble is owed nothing.
The design has four parts:
1. An audit. Every jurisdiction that adopts the Mandate publishes, each year, an account of the Trimunia: the state of its soils, rivers, forests, air and species; what it is bequeathing to the unborn in debt, housing, carbon and depleted stocks; and the condition of mutual care, from loneliness and inequality to who can reach a doctor or a court. That last measure matters more than it first appears. Women perform about three-quarters of all unpaid care in the world, by the International Labour Organization’s count, labour the national accounts leave out entirely. An audit of the living makes it count. An office independent of the executive compiles it, governed by a rotating board drawn by lot. Wales came partway in 2015, when its Well-being of Future Generations Act created a commissioner for those not yet born. The commissioner can advise, review and embarrass, but can’t compel.
2. A chamber. A standing citizens’ chamber, drawn by lot from everyone aged sixteen and over, rotated every eighteen months or so, paid properly and served by its own secretariat. Because its members are drawn by lot and balanced like a census, half of them will be women from the first sitting, a parity that a century of elections has delivered in only a handful of countries. Those too young to sit still have a voice. A children’s assembly, drawn the same way from among the young, meets alongside the chamber, which must hear its findings before issuing any warning or veto. A design that claims to answer to the generations not yet born can hardly ignore the ones already here.
From the model eastern Belgium adopted in 2019 the chamber inherits the power to set agendas and pursue follow-up. To those it adds two powers no deliberative body has yet held. The first is a suspensive veto: legislation the audit shows will degrade the Trimunia can be held back, and the elected house may override it only after a fixed delay and by a supermajority. The second is the power China never built. When the audit records sustained decline across successive years, and the government has ignored the chamber’s formal warnings, the chamber may dissolve the government and send the electorate back to the polls. Heaven’s verdict, delivered by a nurse, a fitter, a student and a woman who sells fruit at the bus station, without a dynasty having to fall first.
3. Clean money and open information, pursued together, since neither reform succeeds without the other. Campaigns are publicly financed and private donations to parties and candidates prohibited. Ownership of news outlets and platforms is capped. Public-interest journalism is funded by a levy on platform advertising and distributed by juries drawn by lot, so that neither ministers nor advertisers decide which reporting survives.
4. Entrenchment. The audit and the chamber are written into the constitution in a form that only the chamber, together with a referendum, can amend. France has already shown why: a reform that depends on the goodwill of those it constrains lasts exactly as long as that goodwill.
Elections remain. Representatives still draft, administer and govern. What fades is the adversarial theatre, because scrutiny now sits with a body that has no party to defend.
For the expert who doubts
You’ve read the deliberative literature, and you know its sharpest critic. The philosopher Cristina Lafont argues that bodies drawn by lot face a dilemma. If their recommendations must still win over the wider public, they add little. If they bypass public deliberation and feed straight into law, they ask citizens to defer blindly to strangers, which corrodes the very legitimacy they were meant to restore. It’s a serious objection, and much of the enthusiasm for citizens’ assemblies has never fully answered it.
The Mandate for Life is built around it. The chamber makes no laws. Its veto delays, and the elected house can override it. Its gravest power, forfeiture, decides nothing at all: it returns the question to the whole electorate. Nobody is asked to defer to anyone. The chamber’s task is to say, with the audit in hand, that the people should be asked again. Think of it as an alarm wired to the ballot box.
The audit can be gamed, you’ll say. Any figure that can unseat a government will attract people determined to bend it, and China’s own record of inflated provincial statistics shows how quickly a target becomes a performance. Two features blunt this. The audit’s board rotates by lot, so there’s no permanent office to capture. Its measures are plural, so improving one at the expense of the others gains nothing. Neither feature makes manipulation impossible; both make it expensive, which is more than can be said of an arrangement where the figures that matter are quarterly growth and next week’s poll.
Scale is the next worry. Eastern Belgium serves fewer than 80,000 people, and Zeguo was a township. Yet Ireland drew national samples to resolve questions that had paralysed its parliament for a generation, and France’s convention spoke for a country of 67 million. Random selection is unusual among democratic instruments in that its accuracy depends on the size of the sample, not on the size of the population from which it is drawn.
Then competence. Wherever lotteries have been given proper conditions (enough time, balanced evidence, moderators who protect the slow and the shy, a real decision waiting at the end), people drawn at random have reasoned well. Zeguo’s sample included residents who couldn’t read, and its judgement still moved coherently away from what officials had assumed. Decades of advertising-funded news and schooling built around employability have deskilled us. But that describes what people do with what they are handed. Capacity follows conditions.
The objection I’d press hardest, though, is one you’re unlikely to raise against yourself. Assemblies are briefed by experts, and most expert briefings on housing, energy, welfare or tax assume the future will be paid for by growth. Every state built on industrial economism needs growth to function; its debts, its pensions and its tax receipts are all calculated on the expectation of more. Seat a perfectly fair lottery inside that assumption and it will deliberate, fairly, over how to share the proceeds of extraction. The procedure will be impeccable and the river will still die. The audit exists to put a second set of numbers on the table, the condition of the living, so that no briefing can pretend growth is neutral ground.
You’ll also remind me that democracies can move fast when they must. They converted whole economies for war, produced vaccines within a year of a new virus appearing, and agreed in Montreal to phase out the chemicals thinning the ozone layer. All true, and revealing. Each time the threat was vivid, near and undeniable. Democracies fail on slow threats because nothing in their calendar makes a falling aquifer or a vanishing species feel like an emergency until it can no longer be met cheaply. The audit exists to do exactly that, putting the slow catastrophe in front of the public every year, with a body empowered to act on it.
I am conscious that three problems stay open. A chamber that seeks the widest agreement can press the stubborn dissenter into silence, so its rules need protected minority reports and a public right of reply. The audit’s methods will need years of contest before they settle. And the Mandate is domestic. Even a nation that governed itself flawlessly would solve only its own share of a planetary commons, and no reform inside one border supplies enforcement across many; Belém showed what happens when nearly two hundred vetoes meet. What the Mandate can change is who arrives at the table, since governments answerable at home for the condition of the living would bargain differently abroad. No such difficulty refutes the design. The present arrangement has graver flaws, and calls them stability.
For the member who holds a seat
I won’t tell you the system is broken. It selected you, rewarded you and taught you how it works, and you have every reason to defend it. So let me argue from your interests instead.
In late 2023 the OECD asked people in thirty of its member countries how far they trusted their national government. Thirty-nine per cent said they did; forty-four per cent said they had little or no trust at all. Political parties fared worse, trusted by about one person in four. The same survey found the strongest predictor hiding in plain sight. Among people who felt the system gave them a say, a large majority trusted their government. Among those who felt it didn’t, barely one in five did. Only about three in ten felt they had any say. Every government standing on that ground is governing on borrowed time, yours included.
The alternatives to sharing power are already available. A legitimate grievance left unanswered doesn’t fade; it finds a louder and more insistent vehicle, and each insurgent who wins and disappoints clears the way for a more extreme successor, less bound by the rules the last one still observed. You can spend your career shoring up your seat against that rising water, or you can stop it at the source by giving people what they say they lack.
Irish politicians learned this on abortion. For decades the issue split parties from within and froze the parliament; a minor reform in 2013 had torn Fine Gael. A Citizens’ Assembly of 99 people broke the paralysis. Critics called it a dodge, a way for ministers to hand a hot question to strangers. Perhaps it was. But it also worked, and the 2018 referendum gave the government a mandate no party could have manufactured alone.
The chamber’s forfeiture power will frighten you most. Look at what it actually does: it sends you back to the voters, with the audit in public view, and you may well be returned. Compare the exits you face now, the leadership spill engineered in a corridor, the donor who withdraws their support, the insurgent who wins on rage. Public financing, meanwhile, ends the hours spent courting money, and with them the obligations money leaves behind.
Jiang Zhaohua found that power given away came back as legitimacy, though it came back to the office, never to him personally. That’s the trade. You’ll own less of the authority you exercise. What you keep will be believed. You might even be trusted.
For the citizen who demands more than outrage
You are right that something is badly wrong, and you’ve probably been told so often enough to feel exhausted by it. Outrage is easy to manufacture and to find. Structural renewal is rarer, and two counterfeits are on offer in its place.
The first is the insurgent who promises to drive out the outsiders, restoring your lost sense of belonging by shrinking the circle of who belongs in the first place. That cure feels like solidarity, because nothing binds people faster than a shared enemy. It works by mining what remains of the common ground until there’s only quicksand left upon which to stand. The second counterfeit is gentler and almost as corrosive: the citizens’ assembly convened as a photo opportunity, thanked warmly, then filtered and set aside. France showed how that ends. Each one teaches people that taking part is a waste of their weekends.
I learned the harder lesson by failing. With colleagues in Melbourne I helped build MiVote, a platform designed to let citizens deliberate on the issues directly concerning them and bind their representatives to the result. Former ministers, including a former Prime Minister, tried to break the design and couldn’t. It foundered anyway, because what was missing lay beneath any design: a sense among citizens that they shared the same problems and a common stake in solving them. We had built a bridge for two riverbanks that had already crumbled. A standing chamber drawn by lot answers that in a way no platform can. It seats a cleaner beside a surgeon, a retired dairy farmer beside a nineteen-year-old who has never voted, for weeks, facing a problem none of them can solve alone. Rotate that through a population for ten years and you have grown a civic habit the present arrangement has spent decades eroding.
So ask for the specific, never the general. Ask your council to bind itself, as a Chinese township once did, to a lottery for one line of its capital budget, and to say in advance that it will fund what the lottery chooses. Ask that any assembly your government convenes come with a published duty to respond, with reasons for every rejection, the way eastern Belgium now requires. When the letter arrives inviting you to serve, say yes. Back the dull reforms that starve the two captures, the donation bans and the ownership caps, because without them every chamber deliberates inside a distorted picture of the world. Then ask every candidate who wants your vote whether they would sign the Mandate for Life, and notice who changes the subject.
The first time
Picture the first time a chamber uses its power. Somewhere modest, a district of a few hundred thousand people, the audit shows the aquifer falling for a fourth consecutive year and child poverty rising, while the council’s answers to three formal warnings amount to a press release. Twenty-four strangers vote to send the district back to the polls. There are lawyers by morning, and columns by afternoon on the tyranny of the unelected, written by people who never once objected to the tyranny of the donated. The mayor campaigns on a pledge to mend the aquifer and may well be returned.
What nobody can predict is what the losing side does next. It may accept that twenty-four strangers spoke for the realm. Or it may reach, as threatened power always has, for whatever lies closest: a compliant judge, a friendly newspaper or talk show host, a general who owes it a favour. On that first morning the Mandate for Life faces the test the Mandate of Heaven so rarely passed in three thousand years, whether a people can take back what it lent without anyone having to burn the palace down.