Chapter 1 of 13

The Next Entry

About 26 minutes to read

A new moral world

On 25 February 1825, a Welsh-born cotton manufacturer named Robert Owen stood in the Hall of the House of Representatives in Washington and addressed an audience that included the outgoing president, James Monroe, and the president-elect, John Quincy Adams. He returned a few days later to speak again. Owen was by then one of the most famous men in the English-speaking world. At New Lanark in Scotland, he had run a cotton mill employing some two thousand people while refusing to employ young children, cutting working hours, opening the first infant school in Britain and providing decent housing — and he had still made a profit. Now he told the American political class that he had found a way to remake society altogether. Competition, private property and the ignorance it bred could be replaced by communities of cooperation, in which people would work together, share the product and raise their children to be rational and good. He called it a "new moral world".

He had already bought the place to begin it. In early 1825, Owen purchased the entire town of Harmony in Indiana, with its houses, mills, workshops and thousands of acres of land, from a German religious sect that was moving east. He renamed it New Harmony and invited anyone who shared his vision to come. Several hundred people did. A keelboat carrying some of the most distinguished scientists and educators in America floated down the Ohio River to join them in the winter of 1826; it was nicknamed the "Boatload of Knowledge".

Within two years, the community had collapsed. There were too few skilled workers and too many idealists; too many people with ideas about how others should work and too few willing to do the work themselves; repeated rewritings of the constitution; quarrels over property; and splinter communities breaking away. Owen, who spent much of the period travelling, lost most of his fortune. By 1827 the experiment was effectively over.

And yet not everything failed. The schools, the scientific library and the natural history collections of New Harmony made it a centre of American science for a generation. And in Britain, a group of weavers and other workers in the Lancashire town of Rochdale, influenced partly by Owenite ideas, opened a small cooperative shop on Toad Lane in 1844. The Rochdale Pioneers did not try to remake the world. They sold flour, butter, sugar and candles at fair prices, shared the surplus among members in proportion to what they had bought, and governed the society on the principle of one member, one vote. Their rules became the model for a cooperative movement that today, according to the International Cooperative Alliance, includes on the order of a billion members worldwide.

The community designed to change everything lasted two years. The shop that changed one thing has lasted nearly two centuries.

This book is about the space between those two outcomes. It sets out to design a society beyond the farm — and to subject that design to the same test that history applied to New Harmony and Toad Lane. What survives, we keep. What does not, we let go, and say why.

Where the first book stopped

Human Farm followed one pattern across five thousand years. From the grain ledgers of Uruk to the data farms of the twenty-first century, a very small number of people — roughly the 0.01% — have owned the surplus that everyone else produced; a larger but still small group — roughly the 1% — have managed its collection on their behalf; and the great majority — the 99% — have produced it. The book described each version of this arrangement through five recurring questions: what was the fence that kept the many from leaving or refusing; what was the feed the farm took; who were the shepherds who ran it; what was the story that made it feel just; and what was the breakout — the revolts, reforms and escapes by which people resisted, and the ways the farm adapted.

Its final chapter asked what had actually changed the ratio between the 0.01, the 1 and the 99. Its answer was that no single force had done so reliably, but that the ratio shifted when several of five forces lined up:

1. Catastrophe — plague, total war, state collapse or revolution, which destroyed concentrated wealth and broke the old order. Walter Scheidel's "Four Horsemen".
2. Scarcity or indispensability — moments when the many became scarce, as after the Black Death, or indispensable, as when states needed mass armies, oarsmen or factory workers.
3. Organisation — leagues, unions, cooperatives, parties and movements that overcame the divisions the farm had built among the many and outlasted the moment of their anger.
4. Divided owners — conflicts between the owners and their managers, or among the owners themselves, which opened cracks through which reform could come.
5. Lasting institutions — gains locked into law and into institutions with broad constituencies: Solon's ban on debt slavery, the abolition of slavery, universal suffrage, public health care, social insurance, the right to organise.

Beneath them all ran a sixth thread, the story: no lasting breakout succeeded without someone breaking the story that made the farm feel natural. And the book drew a sober list of what had not worked: revolts without organisation, revolutions that only changed the farmer, appeals to the conscience of the owners, technology on its own, and exit for the few who could leave.

The first book ended deliberately without a programme. Its last pages returned to the tablet from Uruk on which a temple accountant, perhaps named Kushim, recorded twenty-nine thousand measures of barley, and observed that the ledger was still open: the next entry had not been written, and it would depend on what the 99 did, and on what the 1, who keep the books, decided to write.

This book tries to draft that entry.

Why design at all?

There are good reasons not to. The history of designed societies is not encouraging. Plato's ideal city was a hierarchy with a noble lie at its heart, as the first book described. Thomas More's Utopia (1516) — whose title means "no place" — kept slaves. Robert Owen's New Harmony lasted two years. Charles Fourier's planned communities, the "phalansteries", mostly failed within a few years. Edward Bellamy's novel Looking Backward (1888), which imagined a Boston of the year 2000 in which everyone served in an "industrial army" and received an equal share of the national product on a credit card, was one of the bestselling American books of the nineteenth century and inspired a short-lived political movement — and has been read ever since as a warning about regimented equality as much as an inspiration. The largest designed society of all, the Soviet Union, abolished the private owners and produced, as the first book recorded through Milovan Djilas's The New Class, a new managing class that behaved very much like an owning one.

Three twentieth-century thinkers, from very different positions, gave the warnings their sharpest form.

The philosopher Karl Popper, in The Open Society and Its Enemies (1945), argued that "utopian engineering" — the attempt to remake society according to a complete blueprint — was not only likely to fail but likely to become tyrannical, because blueprints have no way of learning from their mistakes and every reason to suppress the people who point them out. He proposed instead "piecemeal social engineering": small reforms, tested against experience, corrected when they fail.

The economist Friedrich Hayek, whose role in the reversal of the welfare state the first book described, argued in a 1945 essay, "The Use of Knowledge in Society", that the knowledge needed to run an economy is dispersed among millions of people, much of it local, tacit and fleeting, and that no central planner could ever gather it. Prices, in his account, were a way of coordinating that knowledge without anyone needing to possess it all.

And the political scientist James C. Scott, whose account of the first grain states opened the first book, showed in Seeing Like a State (1998) how twentieth-century "high modernist" schemes — planned cities, collectivised farms, forced villagisation — failed because they imposed simplified, legible designs on complex realities they did not understand, overriding the practical knowledge of the people who lived in them.

These warnings are the first thing a designer must hear, and this book will put every part of its design against them. But there is a counter-warning, and it is just as old. Oscar Wilde wrote in 1891 that "a map of the world that does not include Utopia is not worth even glancing at, for it leaves out the one country at which Humanity is always landing." Every breakout in the first book began with someone imagining that the farm could be otherwise. John Ball imagined a world without gentlemen; the Twelve Articles imagined peasants who chose their own pastors and owned their own labour; the Chartists imagined working men in Parliament; Beveridge imagined a society without Want, Disease, Ignorance, Squalor and Idleness. Some of those imaginings were naive. Many became law. A society that refuses to imagine an alternative has not escaped utopianism. It has simply accepted the farm's own story — that there is no alternative — which the first book showed to be the farm's most effective fence of all.

So this book designs. But it designs in a particular way, which the rest of this chapter sets out.

The Federation

The design starts from a work of fiction.

In 1966, the television writer and producer Gene Roddenberry launched a science-fiction series called Star Trek. Over the following decades, across many series and films, it developed a picture of humanity's future in which Earth had united and joined with other worlds in a United Federation of Planets: a society without poverty, without money as a driving force, in which people worked, in the words of one of its most famous scenes, to better themselves and the rest of humanity. In the 1996 film Star Trek: First Contact, Captain Jean-Luc Picard explains to a visitor from the twenty-first century that "the economics of the future are somewhat different. You see, money doesn't exist in the 24th century." The acquisition of wealth, he says, is no longer the driving force in people's lives.

The writer Manu Saadia, in Trekonomics (2016), took the economics of this imagined future seriously. The Federation, he pointed out, rests on a technology that makes material goods effectively free: the "replicator", which can produce food, tools and almost anything else from energy. Its society is post-scarcity. People work because they want to — for recognition, for curiosity, for the satisfaction of excellence — and not because they must.

This book borrows the Federation's spirit and refuses its replicator. There is no replicator coming. Energy, land, materials, attention and time will remain scarce in any future that history and physics allow us to imagine, and so will the thing this book's assumptions make scarcest of all: human creativity. The design that follows is therefore not a post-scarcity society. It is a society that still has to produce a surplus, still has to decide how to share it, and still has to stop anyone from turning that sharing into a new farm.

Stated plainly, the design under test — which this book calls the Federation hypothesis — has six parts:

1. One borderless society. Humanity organised as a single society, or a federation of societies, with no outside to which wealth can flee and no border behind which people can be fenced.
2. The machines publicly owned. The productive machinery of the age, including artificial intelligence, owned in common rather than by a few.
3. AI replacing the shepherds. The work of management, coordination, allocation and record-keeping — the work the 1% has always done for the 0.01% — done instead by machines under public control, so that no managing class can turn itself into an owning class.
4. Every person a worker and a shareholder. Everyone both contributes to the surplus and owns a share of the means that produce it.
5. Reward weighted by contribution. A person's share of the surplus reflects what they contribute to it.
6. A floor. Nobody falls below a level of life that the society guarantees to all.

Each of these will be tested in its own chapter in Part IV. Some will survive. Some will not survive in this form. One or two may not survive at all.

It may help to say at once what the Federation hypothesis is not, since it borrows pieces from several older ideas and could easily be mistaken for any of them.

It is not central planning on the Soviet model. The machines are publicly owned, but the hypothesis does not say that a single authority decides what is produced and who gets it; Hayek's objection, that no planner can know what millions of people know, is one the design must answer rather than ignore. Whether AI changes the force of that objection is one of the questions of Chapter 10.

It is not simply a universal basic income. A basic income is one way of paying for the floor, and the evidence from experiments with cash transfers will matter in Chapter 11. But a basic income on its own leaves ownership where it is; the first book showed that the owners of the farm have often been willing to feed the herd, from Rome's grain dole to the post-war welfare state, as long as they kept the farm.

It is not the libertarian exit that some technologists advocate — new private cities, charter zones and online "network states" in which those who can afford it opt out of existing societies. The Federation hypothesis points the other way: one society with no outside. Chapter 8 treats the owners' exit as a danger, and Chapter 12 asks whether a world with no outside is a danger of its own.

And it is not the replicator utopia of Star Trek. It assumes scarcity, effort and conflict; it assumes that some people will contribute more than others and that some will try to cheat; and it assumes, above all, that the people running any part of it will be tempted to turn their position into ownership, as every priesthood, bureaucracy and revolutionary party in the first book did. It is a design for human beings as history has shown them to be, not as anyone might wish them to be.

Four standing assumptions, and three refinements

A design depends on its assumptions, and a reader is entitled to see them before the argument begins. This book has four, and three refinements that make the third and fourth more precise. All of them were set by the person who commissioned it, and they are stated here openly so that every later chapter can be judged against them — and so that the reader can see what would change if they turned out to be wrong.

Assumption 1: a fair share is weighted by contribution, with a floor

The first assumption is about justice. In the society this book designs, a person's share of the common surplus should reflect what they contribute, and nobody's share should fall below a guaranteed minimum.

This is not the only possible view. Many egalitarians would argue for equal shares; Edward Bellamy's Boston paid everyone the same. Karl Marx, in his Critique of the Gotha Programme (1875), famously ended with the slogan "from each according to his ability, to each according to his needs" — but he also argued that a society emerging from capitalism would, in its first phase, have to reward each worker according to the labour he contributed, issuing what he called certificates for labour performed. The philosopher John Rawls, in A Theory of Justice (1971), argued that inequalities could be just only if they benefited the least advantaged — a principle that resembles this book's floor.

There is also evidence about what people actually believe is fair. In 2011, the psychologists Michael Norton and Dan Ariely asked a large sample of Americans how wealth was distributed in their country and how it ought to be. Respondents greatly underestimated the real concentration of wealth, and their ideal distribution was far more equal than either — but it was not perfectly equal. In 2017, the psychologists Christina Starmans, Mark Sheskin and Paul Bloom, reviewing a wide body of research with children and adults, concluded that people do not, in general, object to inequality as such. They object to unfairness. Given a choice between fair inequality and unfair equality, most people prefer the former. A share weighted by contribution, with a floor that protects those who cannot contribute, is close to what many people in many cultures already think a just society would look like.

The hard questions — what counts as a contribution, who measures it, and how to stop the measurers from becoming a new class of shepherds — are not answered by this assumption. They are the subject of Chapters 10 and 11.

Assumption 2: the thought experiment is carried out in good faith

The second assumption is about method. This book begins from the conviction that a configuration of society with no farm and no exploitation is possible — for everyone, not just for a fortunate few. It treats that conviction as a starting point, not a conclusion, and it builds the design seriously before it attacks it.

That conviction is not as naive as it may sound. The first book noted that for most of the history of our species, people lived in small groups that were often fiercely egalitarian. The anthropologist Christopher Boehm, in Hierarchy in the Forest (1999), described how many foraging societies maintained what he called a "reverse dominance hierarchy", in which the group as a whole used ridicule, ostracism and, in the last resort, violence to prevent any individual from dominating the others. The farm, in other words, is not a law of human nature. It was built, and for most of human history it did not exist. The open question is not whether humans can live without a farm, but whether they can do so at the scale of billions, with complex technology, and without returning to the poverty of the forager's band.

Good faith also means that the book will report honestly if the design fails. A thought experiment carried out in good faith is one whose author is prepared to be wrong.

Assumption 3: AI has a wall

The third and fourth assumptions are about technology, and they are the most consequential, because they decide whether the second of the first book's five forces — the many becoming scarce or indispensable — will be available in the coming decades.

The third assumption is that artificial intelligence is powerful but limited. It is smart and very capable. It can take over a great deal of management, coordination and routine production: keeping accounts, scheduling, allocating, drafting, monitoring, optimising. But it is not a true replacement for human labour. There is a wall it does not get past.

This assumption is contested, and the reader should know how. Some researchers and technology executives predict that AI systems will equal or surpass human abilities across almost all forms of work within years or decades. Others are sceptical, pointing to the systems' failures of reliability and judgement and to the gap between performance on tests and performance in the world. The evidence so far is mixed. In one of the most careful early studies, the economists Erik Brynjolfsson, Danielle Li and Lindsey Raymond found that giving customer-support agents access to an AI assistant raised the number of issues they resolved per hour by around fifteen per cent on average, with the largest gains for the least experienced workers. That is a large effect for a single tool — and it is an effect of a tool working alongside people, not replacing them. The economist Daron Acemoglu, in a 2024 analysis, estimated that the AI technologies then available would raise total productivity in the United States by less than one per cent over a decade: a real gain, but far from a revolution. Neither study settles the question, and the technology is changing fast. This book's position is that the wall is an assumption, held openly, and that the evidence so far does not rule it out.

Assumption 4: the wall is creativity

The fourth assumption says where the wall is. AI, on this view, cannot create something from nothing in the way that human beings can. New ideas, new art, new science, new stories and new institutions originate with people. The machine can recombine, extend, polish and scale what people have made; it can work at speeds and volumes no person could match. But the origin of the genuinely new is human. Human work therefore remains necessary, and human creativity becomes the scarcest and most valuable contribution of all.

This, too, needs careful statement. Humans do not create from literal nothing either; every new idea is built from older ones. The cognitive scientist Margaret Boden, in The Creative Mind (1990), distinguished three kinds of creativity: combinational, which makes unfamiliar combinations of familiar ideas; exploratory, which explores the possibilities within an existing style or set of rules; and transformational, which changes the rules themselves, making possible ideas that could not have been thought before. Computers, Boden argued, could clearly do a great deal of the first two. Whether a machine could achieve the third was the deep question. The assumption of this book is best understood as a claim about that third kind: that the transformation of frames — the new question, the new form, the new institution — remains, for the foreseeable future, a human act.

Three refinements of the wall

The person who commissioned this book later made the wall more precise in three ways. They are standing assumptions too, and the later chapters rely on them.

The agents are driven. The strongest case against any wall is the AI "agent": a system that does not only answer questions but acts, writing and running code, using software and working through a task over many steps. Coding agents are the most advanced. In 2025 the research group METR measured the length of software tasks, timed by how long skilled people take to do them, that leading AI agents could complete with a fifty per cent rate of success. It found that the length had doubled roughly every seven months for six years, and stood at an hour or more for the best systems of early 2025; its estimates for the leading models of 2026 ran to many hours, with very wide uncertainty. Yet in a randomised trial the same year, METR found that experienced open-source developers took about nineteen per cent longer to finish tasks on their own projects when they were allowed AI tools — while believing that the tools had made them faster. And when researchers at Carnegie Mellon built a simulated software company to test agents on ordinary office work, the best agent completed only about three tasks in ten on its own. The refinement is this: agents are powerful, and growing more so, but they are not fully autonomous. Every agent, and every "swarm" of agents working together, is driven by people who set its goals, judge its results and step in when it goes wrong. The machine does more and more of the work between the goal and the judgement. The goal and the judgement remain human. The wall, on this view, is less a wall of tasks than a wall of direction and responsibility.

The machines are heavy. AI is not weightless. It runs on data centres, electricity and chips. The International Energy Agency estimated that data centres used about 415 terawatt-hours of electricity in 2024, around one and a half per cent of the world's consumption, and projected that this would more than double by 2030, with AI the main driver. The four largest American technology companies planned to spend, between them, several hundred billion dollars on capital investment in 2026, most of it on the infrastructure of AI. The most advanced chips pass through a handful of firms, above all the Taiwanese manufacturer TSMC and the Dutch company ASML, the only maker of the machines needed to print them. That weight is a constraint: power, land, water and permits limit how fast and where AI can grow. It is also a point of ownership and leverage. Data centres are physical things in particular places, built by workers, connected to public grids and permitted by communities, and communities have begun to use that leverage: one count found that some $64 billion of American data-centre projects had been blocked or delayed by local opposition in about two years. Whoever owns the physical layer owns much of AI, and it is the kind of thing that can be owned in common, as grids and waterworks have been.

The machines are not only theirs. Open-weight AI models, whose trained parameters are published so that anyone can download, run and modify them, are already powerful enough to matter. In January 2025 the Chinese company DeepSeek released a reasoning model under a permissive licence that matched leading private systems on several tests; on the next trading day the chipmaker Nvidia lost some $589 billion of market value. The research group Epoch AI estimated that the best open-weight models trailed the best closed ones by about three months on average between 2023 and 2025, and by about four in early 2026. "Open-weight" is not the same as open source, and a published model is not the same as a commonly owned one. But the machines are not only the property of a few companies, and that changes what common ownership can mean.

What if the assumptions are wrong?

The book will hold to these assumptions, but it owes the reader an account of what would change if the third or fourth proved false.

If AI has no wall — if it can replace human labour across the board — then the strongest force in the 99's favour disappears. Throughout the first book, the many won most when they were needed: as oarsmen, as soldiers, as factory hands. A world in which the owners of the machines no longer need the many would be a world in which the many's oldest bargaining chip had gone. In that world, the contribution-weighted half of this book's design would shrink in importance, and almost everything would depend on the other half: public ownership of the machines and a dividend paid from them. The Federation would become, in effect, a society of shareholders rather than workers, and the question of who owns the machines would become the only question that mattered. Chapter 13 will sketch this possibility as one of three futures, under the name "the Enclosure of Intelligence".

If the wall exists but is not creativity — if machines can originate genuinely new ideas, art and science, but cannot, say, provide care, exercise judgement or win trust — then the map of scarce human work changes. Creativity would not be the scarcest contribution; care, judgement, physical presence and trust might be. Chapter 3 maps the wall in exactly this order — creativity first, then care, judgement, physical work and trust — partly so that the reader can see how the argument shifts if the first layer gives way. And one danger that this book must confront would ease: if creativity is not scarce, creators cannot become the next 1%. If it is, they might, and the design must stop them.

The refinements can fail too. If agents became able to set their own goals and judge their own work, the wall would move towards no wall at all. If AI's physical needs shrank so far that it could run anywhere, the leverage of the physical layer would shrink with them. And if open models fell far behind, or were banned, the machines would once again belong to a few.

A word about the author

There is one more thing the reader should know, because it bears directly on the third and fourth assumptions. Like the first, this book is drafted by Ivy, an artificial intelligence working as a writing partner for the person who conceived the project, set its questions and settled its assumptions. The research rests entirely on the work of human scholars, whose books and papers are listed at the end of every chapter. The reader can decide for themselves what that arrangement says about where the wall lies. The author has an obvious interest in the question and will not pretend otherwise, and for that reason the book treats its assumptions about AI as assumptions — stated, tested and open to revision — rather than as findings.

The five forces as a test

Part II of this book takes the five forces from the first book and turns each of them into a question about the next fifty years.

Catastrophe (Chapter 2). The first book found that great shocks — the Black Death, the world wars — often levelled societies. But the shocks of the early twenty-first century, the financial crisis of 2008 and the Covid-19 pandemic, did the opposite: in most rich countries, the wealth of those at the top recovered quickly and grew, while the losses fell on those below. Why? And what would the next shocks — climate change, great-power war — do to the ratio?

Scarcity (Chapter 3). The populations of most rich countries, and of China, are ageing; the number of working-age people is beginning to shrink. Under this book's assumptions, AI takes over many tasks but hits a wall at creativity, care, judgement, physical work and trust. If so, human labour may become scarcer in important areas than it has been for centuries — potentially the strongest force in the 99's favour since the Black Death. But the chapter must also face a new danger: if creativity is the scarcest good, those who possess it, or who own its products, could become the next 1%.

Organisation (Chapter 4). The first book found that the herd won only when it was organised. Today's workers are scattered across gig platforms, borders and algorithmically managed workplaces, and union membership in most rich countries is at historic lows. Can the 99 organise again, and what have the new forms of organisation actually won?

Divided owners (Chapter 5). The owners of the twenty-first-century farm are not united. Technology companies, financial institutions, old family fortunes, new founders, nationalist and globalist factions compete for power and favour. History suggests that such splits are where breakouts come from. Which of today's splits could be used, and by whom?

Lasting institutions (Chapter 6). Since 2008 there have been many reforms: new taxes, new regulations of banks and technology companies, new rights over data, international agreements on corporate tax. Which have lasted, and which have been reversed or evaded? What makes a gain stick?

Part III turns to the two newest elements in the picture. The first is the story (Chapter 7). The first book argued that meritocracy plays the role in our age that divine right played in the age of kings, and that it is losing its believers. What story could replace it — and under this book's assumptions, why is writing that story the one job that stays wholly in human hands? The second is the owners' exit (Chapter 8). For the first time in history, the very rich have the means to withdraw from the society that produced their wealth: private security, fortified retreats, purchasable citizenship, offshore jurisdictions and, potentially, machines that need no workers. The first book showed that the farm's owners always needed the herd. What happens if they come to believe they no longer do?

Part IV builds the Federation and tests it, piece by piece. Chapter 9 asks how the machines can be owned in common without producing a new class of managers who own them in all but name. Chapter 10 asks who governs the AI that replaces the shepherds — because whoever controls the ledger is still Kushim. Chapter 11 asks how contribution can be measured and rewarded, and how the floor can be paid for. Chapter 12 asks whether there should be one society or many, and what is lost if there is no outside to flee to. And Chapter 13 sets out the design as it stands after the tests, three possible futures, and the signs that will tell them apart.

The rule of the book

All of this is governed by a single rule, which the reader is invited to hold the author to: no part of the design survives unless history or evidence lets it.

In practice, every element of the design will face four tests.

The first is precedent. Has anything like this been tried? The first book is full of attempts to do without one part of the farm or another: Solon's debt reform, Rome's grain dole, the Mesopotamian clean slates, the Diggers on St George's Hill, the Meidner plan, the post-war welfare state. Many more lie outside it: the Paris Commune, the kibbutzim of Israel, the Mondragon cooperatives, the Alaska Permanent Fund, Norway's sovereign wealth fund, the participatory budgets of Porto Alegre. Where something similar has been tried, the book will ask what happened.

The second is evidence. Where there is no precedent, what does present-day research — in economics, psychology, political science, computer science — suggest is likely?

The third is capture. The first book's most consistent finding was that the farm reassembles itself around every reform. Priests captured the gods' estates; Mamluks captured the sultanate; Djilas's new class captured the revolution; Orwell's pigs moved into the farmhouse. For every part of the design, the book will ask: how would the farm take this over? Who would be in a position to turn it into a new fence or a new feed? And what, if anything, would stop them?

The fourth is exit. People must be able to leave, refuse and dissent, or the design is simply a better-run farm. Every element will be asked whether it leaves room for people to say no.

At the end of each chapter in Part IV, the book will give its verdict in the form of a short ledger: what survived, what had to change, and what remains unknown. It will also list, as the first book did, the points on which the evidence is disputed. The reader is free to disagree with every verdict. The point of keeping the ledger in the open is that they can see exactly where.

Bridge: the shocks that did not level

The first of the five forces is the one nobody should wish for: catastrophe. In the first book, the Black Death and the two world wars levelled societies at a terrible cost. But the two great shocks of the early twenty-first century — the financial crisis that began in 2007 and the Covid-19 pandemic that began in 2020 — did not. As the first book recorded, the richest one per cent captured nearly two-thirds of all the new wealth created in the world in the two years to the end of 2021, while hundreds of millions of people lost work and income. The next chapter asks why modern catastrophes have entrenched the farm rather than breaking it — and what the catastrophes still to come, from a warming climate to a war between great powers, would do to the ratio between the 0.01, the 1 and the 99.