Everyone gets a B+. Congratulations.

Part three of a three-part series on incentives, capital allocation, and the machine being built to fix both. Part one showed how the West pays for the wrong things. Part two showed who allocates the capital and how badly. This part confronts the ideology now gaining ground on both sides of the Atlantic: that the fix for our problems is to flatten the rewards altogether.

There’s a story the internet has loved for at least fifteen years. An economics professor’s students insist that socialism works, that nobody should fail and nobody should excel. So he proposes an experiment: all grades will be averaged, and everyone will receive the same mark. The first exam averages a B; the students who studied hard are annoyed, the ones who didn’t are delighted. So less studying happens. The second exam averages a D. By the final exam the class fails, and the professor delivers the moral: when reward is divorced from effort, effort stops.

Let’s be straight about this story: it never happened. It is a parable, not a case study. Here is the interesting part: the reason the story refuses to die is that the mechanism it describes is one of the most replicated findings in experimental economics. When participants share the returns of a common pot regardless of individual contribution, experiments and experience show that contributions fall round after round toward near-total free-riding. Every time, in every population tested. People start out generous. Then they notice the shirkers eating the same dinner, and by the final rounds almost nobody brings anything to the table. The professor is fictional. His classroom is running constantly, in laboratories all over the world, with real money, and it always ends the same way.

Swap the students for adults and the grades for money, and you have the live political question of the 2020s. Because a growing and increasingly successful movement in Britain and America looks at the wreckage described in the first two essays of this series, correctly observes that something is badly wrong, and prescribes the classroom experiment, at national scale, as the cure.

In Britain, the Green Party elected Zack Polanski as leader in September 2025 and something unusual happened: the party surged. Membership rose by around 80 per cent to over 100,000, and by April 2026 YouGov had the Greens at 18 per cent, ahead of the Labour government itself. Polanski’s signature policy is a wealth tax: 1 per cent annually on assets above £10 million, 2 per cent on billionaires, part of a package he says would raise over £30 billion a year. “It is a political choice,” he says, “to keep children in poverty whilst billionaires and multimillionaires get richer.” The same party has proposed capping the pay of any CEO at ten times their lowest-paid worker, a policy with 65 per cent public support in YouGov polling, because of course it has; almost nobody in that 65 per cent expects to be the one capped.

Alongside him stands Gary Stevenson, the ex-Citibank trader turned YouTube phenomenon, one and a half million subscribers, author of the memoir The Trading Game. Stevenson’s pitch is seductive precisely because it comes wrapped in market credentials: I got rich betting that inequality would destroy the economy, and I’m here to tell you the rich will buy up everything unless we tax fortunes above £10 million. (Honesty compels a footnote: the Financial Times interviewed eight of his former colleagues who disputed his claim to have been Citibank’s most profitable trader in the world, and his macroeconomics has been publicly savaged not just by free-marketeers like Julian Jessop but by wealth tax supporters on the left like Richard Murphy. The messenger is contested. The audience is real.)

In America, Bernie Sanders and Rashida Tlaib reintroduced the Tax Excessive CEO Pay Act in September 2025, escalating corporate tax rates for any company whose CEO earns more than fifty times its median worker. And the demand side is genuine: an IEA-commissioned poll found 67 per cent of young Britons would prefer to live in a socialist economic system, and Gallup found in August 2025 that only 54 per cent of Americans now view capitalism positively, the lowest figure Gallup has ever recorded. Among Democrats, socialism now outpolls capitalism by 24 points.

Before dismantling the programme, pause on a puzzle: why do checkable facts bounce off intelligent people? The answer starts in physiology. Your brain is about 2 per cent of your body weight and burns roughly 20 per cent of your resting energy; deliberate, effortful reasoning is among the most metabolically expensive things you can do with it. So evolution fitted us with what Daniel Kahneman called System 1: fast, cheap, intuitive judgment that reaches the conclusion first and lets the slow, glucose-hungry machinery stay in bed. We are not truth-seeking engines that occasionally get lazy. We are energy-conservation engines that occasionally, reluctantly, and at real caloric cost, do some thinking. Believing whatever your tribe believes is free. Checking is expensive. Most people, most of the time, quite rationally don’t pay.

Here is the properly unsettling part: intelligence doesn’t fix this, it arms it. Yale researchers gave people a slightly tricky statistics problem. Framed as data about a skin cream, the numerate solved it and the innumerate didn’t, exactly as you’d expect. Framed as identical data about gun control, the most numerate participants became the most likely to get it wrong, whenever the correct answer hurt their political side. More brainpower meant more sophisticated ways to arrive at the conclusion the tribe required. They called it motivated numeracy. Upton Sinclair had already called it, eighty years earlier and in eleven words: it is difficult to get a man to understand something when his salary depends upon his not understanding it.

Britain staged the live demonstration in March 2025, when Stevenson debated the entrepreneur Daniel Priestley on The Diary of a CEO, a debate watched by millions and settled, on the evidence, by nobody. Priestley emerged from the studio and posted a thread that went viral, opening: “He didn’t share any facts. Not one. He wants us to accept his entire philosophy on feels.” And yet the view counts climbed, the book kept selling, and not one mind visibly changed, because nobody in that audience was buying facts. They were buying confirmation, which is cheaper, and comes with a story about whose fault everything is.

And notice that the incentive rule of this series applies to the leaders as much as the led. Stevenson’s one and a half million subscribers arrived for one story and would leave on its retraction; Polanski’s surge is collateralised entirely by the wealth tax. Neither man needs to be lying, and I’m not saying they are. Sinclair’s point is worse than that: a salary that depends on not understanding produces sincere, fluent, unbudgeable not-understanding, immune to any fact you can bring, because the fact is competing with the mortgage. This is why debating these men feels like playing chess against an opponent who is paid per draw. Daniel Priestley can’t win the argument. You can’t win the argument. Nobody can win the argument.

There is no argument. There is an audience. There is a revenue model.

Concede the diagnosis its due. Asset prices have run away from wages; a young worker doing everything right cannot buy the house their parents bought on one salary; and the first essay in this series documented, in some detail, that the game is genuinely rigged in places. The levellers’ voters are not fools. They are people who looked at a broken ladder and concluded the ladder was the problem. And checking whether the ladder is actually the problem costs glucose that the slogan does not.

So check. Stevenson’s central image, the rich buying up everything until they own it all, is a testable claim, and Britain has been keeping the test data since Queen Victoria. In 1895 the richest 1 per cent owned about 70 per cent of the country’s wealth. By 1980 that share had collapsed to roughly a fifth, and there it has sat for forty-five years, through Thatcher, Big Bang, the financial crisis, quantitative easing and the entire YouTube career of Gary Stevenson, barely moving. The century of runaway plutocracy appears in the data as its exact opposite: the greatest dispersal of wealth in British history, followed by four decades of flat. The houses really are unaffordable and the young really are stuck, and those have real causes: planning law, money printing, the incentive wreckage of part one. “The 1 per cent are swallowing everything” is not among them. One chart is all it takes, and the whole point of the previous section is that nobody spends the glucose to look at it.

But now examine the cure, because the cure is the classroom experiment, and the classroom experiment has been run. Not in theory. In countries. Repeatedly. At every scale from a village to a superpower, with results among the most consistent in the entire historical record.

Start with the cleanest one, because it is the least known and the most devastating: the Israeli kibbutz. For most of a century, hundreds of kibbutzim ran full income equality voluntarily. Communities of true believers, self-selected idealists who chose equal sharing and raised their children in it. No Stasi, no gulag, maximum goodwill. Economists spent years studying what happened, and the findings read like a controlled demonstration of every incentive principle in this series. Equal sharing produced systematic brain drain: the ablest members left, because they could capture their wage premium outside and couldn’t inside. It produced adverse selection: applicants to join were disproportionately those with the least to lose from equal pay. And it produced free-riding that the communities policed only through intense social surveillance and by making exit costly. When Israel’s 1980s debt crisis stripped away the subsidies, the kibbutzim faced reality, and by around 2010, roughly three quarters of them had voted to abandon full equal sharing for differential, market-linked salaries. Sit with that. The people who believed in economic equality more sincerely than anyone alive, who built their whole lives around it, abandoned it when they finally had to pay its true cost.

Nobody defeated the kibbutz movement. It graded its own exam.

The involuntary versions were crueller and larger. The Soviet Union permitted rural families tiny private plots alongside the collective fields, and in doing so accidentally ran the twentieth century’s most elegant agricultural experiment: same farmers, same soil, same weather, two incentive systems. The private plots occupied around 3 per cent of cultivated land and produced roughly a quarter of the value of Soviet agricultural output; by 1977, kolkhoz families were getting about three quarters of their meat and eggs from their own holdings. The collective side’s shortfall was blamed, harvest after harvest, on the weather, which in the Soviet Union had apparently learned to stop at the fence. The same man who shirked in the collective field worked his own plot like a demon, because on one side of the fence effort flowed to his family and on the other it dissolved into the average.

China ran the same experiment with eight hundred million subjects. In December 1978, in the starving village of Xiaogang in Anhui province, eighteen households met in secret and signed a document dividing the collective’s land into family plots, pledging to raise each other’s children if the signatories were jailed. One farmer, Yen Jingchang, later explained the result in a sentence that deserves to be carved over the door of every economics faculty: “We all secretly competed. Everyone wanted to produce more than the next person.” The first harvest reportedly exceeded the previous five combined. Deng Xiaoping, rather than jailing them, made their crime national policy. Under the household responsibility system, grain output rose from about 305 million tonnes in 1978 to about 407 million by 1984, and over the following four decades roughly 800 million people exited extreme poverty, three quarters of all the poverty reduction on earth in the period, according to the World Bank. Nothing about the Chinese farmer changed in December 1978 except the answer to one question: if I grow more, do I keep it?

Where the experiment ran in parallel, with a control group, the results were starker still. Germany split one people, one language, one culture, one work ethic down the middle for forty years; when the wall fell, East German productivity was somewhere between 30 and 43 per cent of West German levels, and the East is still catching up a generation later. Korea ran the same split and kept it running: the South now out-produces the North roughly thirty-fold per head, a gap you can see from orbit, in the famous satellite photograph of a black North Korea floating between the electric blazes of Seoul and China. And for anyone who thinks these are Cold War relics, Venezuela performed a modern replication: between 2014 and 2021 its economy contracted by about three quarters, the deepest peacetime collapse ever recorded for a country not at war, on top of the world’s largest proven oil reserves, and nearly eight million people, better than a quarter of the population, walked out.

Kibbutzniks, Soviet peasants, Chinese villagers, undergraduate lab subjects and two halves of the same nation. It is not a right-wing talking point. It is one of the most robust findings in the human record.

Sever reward from contribution, and contribution stops.

“But nobody’s proposing Maoism,” the levellers reply, fairly. “We’re proposing 1 or 2 per cent on fortunes over £10 million.” So look at the modest version’s track record, because Europe has run that experiment too.

In 1990, twelve European countries levied a net wealth tax. Today three do. They abandoned them not in a fit of ideology but because the taxes kept failing in the same ways: brutal to administer, punishing to illiquid founders, and prone to chasing away the exact people who paid the most of everything else. France’s ISF is the textbook case: it raised around €2.6 billion a year, drove an estimated €200 billion of capital abroad since 1988, and cost the state roughly twice its own yield in lost revenue elsewhere. A tax with negative proceeds. Emmanuel Macron, no libertarian, abolished it in 2017. Norway provided the recent replication: when its Labour-led government nudged the top wealth tax rate from 0.85 to 1.1 per cent in 2022, more billionaires and multimillionaires left in one year than in the previous thirteen combined, taking an estimated $54 billion in fortunes, mostly to Switzerland; the country’s third-richest man now allocates his capital from Lugano. Wealth tax receipts actually rose, which its defenders cite, but receipts were never the question. The question is where the next thirty years of investment, company formation and risk-taking happen, and that left in the moving vans.

Britain is already running its own preview. The abolition of non-dom status in April 2025 coincided with a reported exodus of wealthy residents; the widely cited Henley figures, 10,800 millionaires gone in 2024 and a projected 16,500 in 2025, come from a data provider whose methodology has been seriously challenged. But the direction was visible enough that the Chancellor was softening the rules within months. Capital, unlike the tax base the levellers imagine, has legs. A wealth tax on the most mobile people in history is the £200 shoplifting rule from part one, inverted: it posts a price, and the people it prices simply walk out of the shop.

And the 10:1 pay cap is the same error in miniature. Run the arithmetic: in any firm employing a minimum wage worker, the cap sets a ceiling of roughly £250,000 on every salary in the company, investment bankers, chief surgeons of private hospital groups, the lot. The predictable responses, and every one of them is a Hanoi rat, are to outsource the cleaners so the lowest-paid “employee” is a well-paid engineer, to shift pay into options and perks, or to automate the low-wage roles out of existence entirely, at which point the policy has fired the very people it was flattering. There is exactly one sizeable organisation on earth that runs compressed ratios successfully, the Mondragon co-op federation in the Basque country, at roughly 6:1, and its own record is the caveat: it struggles to recruit senior talent, its 122 foreign subsidiaries are conventional firms rather than co-ops, and its flagship consumer business went bankrupt in 2013. Voluntary compression, chosen by members who can leave, is a legitimate niche. Mandated compression is a nationwide experiment in discovering how creative HR departments can be.

Which brings us to the beating heart of the leveller case: the billionaire, hoarding a dragon’s pile while children go hungry. The second essay in this series has already done the heavy lifting here, so let’s just connect the wires.

The moral framing assumes a billionaire’s wealth is a pile of consumption, a Scrooge McDuck vault that could simply be ladled out. Overwhelmingly it is not; it is working capital, equity in operating companies, and the only question that matters for everyone else is who allocates it and how well. We have data on this. Jeff Bezos ran Amazon for two decades at deliberately near-zero profit, recycling every dollar into logistics, capacity and research; Amazon now spends more on R&D, about $86 billion a year, than most governments spend on science in total. Musk’s capital, allocated with his own money at risk on fixed-price terms, built rockets NASA’s own cost model said should have cost ten times more. Compare the state’s showcase allocations from the same essay: HS2, three times the money for less than half the railway; DOGE’s $2 trillion of promised savings dissolving into an unaudited $215 billion claim; and Malaysia’s sovereign wealth fund, 1MDB, which allocated $4.5 billion into superyachts, Picassos and the financing of The Wolf of Wall Street, a film about fraud, paid for with fraud. Britain runs the same machine at pound-shop scale. The chief executive of the Scottish National Party spent twelve years embezzling more than £400,000 of the money the faithful had given for independence, and the police photographs of what he bought with it read like the 1MDB catalogue reshot for Fife: a £124,550 motorhome parked on his mother’s driveway in Dunfermline with four miles on the clock and a £220 designer teapot inside. He is now in prison, and Scotland is no more independent than the motorhome was driven. Every allocation regime gets the 1MDB it can afford; Scotland’s came with a chemical toilet. The billionaire’s capital is disciplined by markets that reprice his decisions every trading day and can take the whole pile away; ask any founder who missed a technology shift. The state’s capital is disciplined by an election every four or five years, fought about everything except capital allocation.

None of this requires believing billionaires are saints. It requires only the observation that a wealth tax is a proposal to move capital, at scale and annually, from the most feedback-disciplined allocators in the economy to the least, the very institutions whose record fills this series. If your complaint is that the rich rig the rules, ferociously prosecute the rigging: the subsidies, the planning capture, the bailouts, the cost-plus contracts. That is an incentives problem, and this series is nothing if not sympathetic to it. But confiscating the working capital of the productive because some of them cheat is averaging the class’s grades because some students copied homework. It punishes the studying, not the cheating, and the second exam comes back a D.

Strip every argument in this essay to its chassis and the same thing is holding it up: the gradient. The slope between doing more and doing less; the gap between what the worker takes home and what the shirker takes home. Steepen it and people climb. Flatten it and they sit down.

Every functioning economic system in history has run on it, and every attempt to abolish it, voluntary or violent, modest or total, has produced the same decay curve as the laboratory: effort collapsing round by round as the able realise they are working for the average.

Britain, as part one showed, has been flattening its gradient from both ends without ever voting on it: a benefits package that outpays 6.2 million workers’ take-home, marginal tax rates above 100 per cent at the £100k cliff, a first rung sawn off the wage ladder, theft de facto decriminalised below £200. The levellers look at this exhausted, rigged flatness and propose, in perfect sincerity, more flattening. The philosopher Michael Sandel says even a perfect meritocracy would be a tyranny, and perhaps he’s right that status games corrode souls. But the practical alternative to allocating rewards by contribution has never been allocation by kindness. It has been allocation by birth, by party card, by patronage, or by queue, and every one of those is crueller to the poor than the gradient ever was. The poorest people in the world’s recent history were made incomparably better off by exactly one force: the restoration of the gradient in China in 1978.

And here, finally, the series closes its loop, because there is now a working system that answers the levellers’ best point and keeps the gradient, and it is the machine from part two. The honest kernel inside the anti-meritocracy case is that the existing ladder is gatekept: the right school, the right accent, the right postcode, the right cap table. Bittensor’s incentive mechanism is what a meritocracy looks like with the gatekeeping physically removed. The network does not know your university, your age, your country or your surname. It cannot see them. A miner is a wallet address and a stream of output, scored by consensus and paid by the block, and the teenager in Jakarta competes on precisely equal terms with the lab in San Francisco. Jacob Steeves, who co-founded the thing, states the design intent plainly: “No one cares about where you went to school. If you can just solve that problem, you get paid out.” No inherited advantage survives contact with a mechanism that pays only for measured contribution, and no bureaucrat can award their nephew a subnet. It is simultaneously the most unequal system imaginable, rewards proportional to value, no floor, no ceiling, no averaging, and the most radically fair one ever built, because the only thing it can measure is what you actually did.

That is the choice on the table, and it is not between compassion and cruelty. It is between two responses to a rigged game. One response flattens the scores, and we have run that experiment from the kibbutz to Pyongyang: the class stops studying, the field goes untilled, the lights go out, and the people it claimed to protect eat last. The other response fixes the scoring, pays ruthlessly and transparently for real contribution, and lets anyone on earth sit the exam.

Everyone gets a B+ sounds like mercy. It is actually a sentence, passed on every child in the class who would have worked, and on every adult in the country who still wants to. The kindest thing you can give the hardworking is a world where the work counts.