Chapter 11 of 13

Contribution and the Dividend

About 23 minutes to read

The day the women stopped

On the morning of Friday 24 October 1975, Iceland stopped working. Banks, shops, schools, fish-processing plants, telephone exchanges and newspapers closed or ran on skeleton crews. Theatres cancelled their performances. Flights were disrupted because the cabin crews were missing. Fathers turned up at their offices with their children, because there was no one else to look after them, and handed them sweets and colouring pencils to keep them quiet; sausages, which could be cooked quickly, sold out in the shops. In the centre of Reykjavík some twenty-five thousand women — in a country of about two hundred and twenty thousand people — filled the square of Lækjartorg and the streets around it to listen to speeches and sing.

The organisers, a coalition of women's groups from the radical Redstockings to conservative housewives' associations, had chosen to call it a "day off" rather than a strike, so that women who feared for their jobs could join. About nine in ten of the women of Iceland took part. They did not go to their paid work, and — this was the point — they did not do the unpaid work either: no cooking, cleaning or childcare. The aim was to show what women contributed by showing what happened when they stopped. The country, it turned out, could not function for a single day without them. The following year the Icelandic parliament passed a law guaranteeing women and men equal status and equal rights. Five years later Iceland elected Vigdís Finnbogadóttir, a divorced single mother, as president, the first woman in the world to be elected head of state in a national vote. Icelandic women have repeated the day off several times since, most recently in October 2023, when the prime minister joined them.

The women of Iceland were making an argument about a ledger. Much of what they did every day was work by any reasonable definition — feeding, cleaning, teaching, nursing, organising — and none of it appeared in the national accounts, the wage statistics or the pension records. It did not count, because no one had decided to count it.

The previous chapter ended with the same problem. The design under test rewards contribution: above a floor that nobody falls below, people receive more for giving more. That means someone — or something — has to decide what counts as contribution and how much it is worth. Whoever decides that decides what people are paid for, which work is honoured and which is ignored. It is the oldest power of Kushim's ledger: not only to record the barley, but to decide what goes into the record. This chapter takes up the fifth and sixth elements of the design — reward weighted by contribution, and a floor that nobody falls below — and asks how they can be built without the ledger becoming a new farm.

What the ledger counts

Modern societies count their production through national accounts, a system invented in the 1930s and 1940s to help governments manage depression and war. Simon Kuznets, who prepared the first official estimates of national income for the United States in 1934, warned Congress in his report that the welfare of a nation could scarcely be inferred from a measure of national income. His warning was largely ignored, and his choices about what to leave out became the world's. Gross domestic product counts what is bought and sold, and the work of government; it leaves out almost everything done without payment inside households and communities. If a man marries his housekeeper and she goes on doing the same work without wages, the old economists' joke runs, the national income falls.

The New Zealand politician and economist Marilyn Waring made that joke into an indictment in If Women Counted (1988). The rules of national accounting, she showed, made most of the work of most of the world's women invisible: the water carried, the food grown for the family, the children raised, the sick nursed. In 2018 the International Labour Organization estimated that people around the world spend some 16.4 billion hours every day on unpaid care work — the equivalent of about two billion people working full time without pay — and that women perform more than three quarters of it. Valued at the minimum wage, the ILO calculated, it would amount to about nine per cent of the world's GDP.

Feminists have argued for half a century about what to do with that discovery, and their argument is a warning for the design. In 1972 the International Wages for Housework Campaign, founded by Selma James, Mariarosa Dalla Costa, Silvia Federici and others, demanded that the state pay women for housework. Their point was partly tactical: to name the work by demanding a wage for it, and so to show that the whole economy rested on it. Other feminists objected that paying for housework would tie women more firmly to the home, turning an unjust division of labour into a salaried one. Both sides were right about something. Counting care is the only way to stop it being invisible; but counting it as one group's natural job would freeze the old division in place. The design must count care as contribution by whoever does it — as Germany's pension credit can be claimed by either parent — and pair it with the time and services that let people move between care, paid work, learning and creation over a lifetime.

The first book found that every farm counted what its owners valued. The temple ledgers of Uruk recorded barley and labour owed; the plantation account books recorded enslaved people as property; the factory's books recorded hours and output, never the household that sent the worker to the gate each morning fed and clothed. Every ledger was a decision about whose work mattered. The design's ledger will be one too. The question is whether it can be made honestly.

What counts

Chapter 7 set the first rule. If reward for contribution is told as a story of desert — that those who receive more are worth more — it will reproduce the cruelties of meritocracy in a new form. The design must justify reward as reciprocity: a fair return for what people give to a common enterprise, and an encouragement to give, while saying openly that talent, health and opportunity are largely matters of luck. The philosopher Michael Sandel's idea of "contributive justice" adds the second rule: people need not only a fair share of income but the recognition of being needed, of contributing something valued to the common good. A narrow definition of contribution would deny that recognition to most people, as meritocracy did.

The plan of this book named six kinds of contribution, and the design should count all of them from the start:

• Production: the making and moving of goods and services, much of it now done with the help of machines.
• Care: raising children, looking after the sick, the disabled and the old, whether paid or unpaid.
• Learning: education and training, which are investments of time and effort that the whole society later draws on.
• Creation: new ideas, art, science, stories and institutions — under this book's assumptions, the work machines cannot do.
• Risk: dangerous, unpleasant or uncertain work, and the willingness to start something that might fail.
• Maintaining the commons: keeping up the shared things — the parks and rivers, the open software and the encyclopaedias, the associations, councils and assemblies that Chapter 10 made central to the governance of the machines.

Counting these does not require inventing a new science. Several societies already count some of them. Germany's pension system credits a parent with three years of contributions, valued as if they had earned the average wage, for every child born since 1992, and has been extending the credit for older children; other European countries give similar credits for raising children or caring for relatives. In Japan in the 1990s, a welfare foundation created fureai kippu, "caring relationship tickets", a system in which volunteers who help elderly people earn credits that they can use later for their own care or transfer to a parent who lives elsewhere. Time banks, popularised in the United States by the lawyer Edgar Cahn from the 1980s, credit one hour of help to a neighbour as one hour, whatever the help was. Ireland's Basic Income for the Arts, a pilot that from 2022 paid two thousand artists €325 a week, treated creative work as worth paying for regardless of whether it sold; in its budget for 2026, the Irish government announced that the scheme would become permanent.

These examples suggest a principle the design can adopt. The main unit of contribution should be time given to recognised work in any of the six fields, with an hour of care counted on the same terms as an hour of production. On top of that equal base, modest premiums could reward work that is especially demanding, dangerous, unpleasant or scarce. That is a deliberate choice against the market's way of valuing work, in which an hour of a financial trader's time can be worth a hundred hours of a care worker's. It is closer to the way a cooperative or a time bank values it. It also has a cost, and the book should say so: it rewards effort and time more than results, and it will be harder to measure some kinds of contribution — above all creation — in hours at all. The creators' question gets its own section below.

Risk needs a word of its own, because the defenders of great fortunes have long argued that they are the reward for it: the entrepreneur who stakes everything on an idea deserves whatever the idea earns. The design takes risk seriously as a contribution, but it changes its terms. Most of the risks that ruin people under the farm are the risks of falling: losing a home, a livelihood or one's health because a venture failed. A floor removes the worst of them, and the evidence suggests that this encourages enterprise rather than smothering it. When France reformed its unemployment insurance in 2002 so that unemployed people who started a business could keep their benefits for a time, the number of new firms rose by about a quarter, according to a study by Johan Hombert and colleagues, and the new firms were not of lower quality. The Kenyan villages that received regular payments, described below, saw people move into businesses of their own. In the design, the person who starts something new is rewarded through the premium for risk and protected by the floor if it fails; what they cannot do is turn a successful venture into a permanent claim on the commons, for the reasons the creators' section sets out.

The measurers

The danger Chapter 7 named is that the measurers of contribution would become a new priesthood, deciding whose work counts. History supplies a sharper warning. When the Soviet Union rewarded output, it got the output it measured rather than the output it needed. Planners set targets in tonnes, and factories made goods too heavy to use; they set them in units, and factories made goods too small. In August 1935 the miner Alexei Stakhanov was reported to have cut more than a hundred tonnes of coal in a single shift, fourteen times his quota; the record, achieved with a team of helpers arranged for the occasion, was used to raise norms for everyone. The British economist Charles Goodhart gave the general rule its name in 1975: any measure that is relied on for control tends to stop measuring what it used to measure. Every scoring system invites people to game it, and the more is at stake, the more they will.

The modern machine makes the danger larger. It could, in principle, measure everything — every hour, every task, every keystroke. Chapter 4 showed what that looks like in the gig economy: workers managed and rated by an algorithm they cannot see. China's "social credit" initiatives, which are in fact a patchwork of local and sectoral schemes rather than the single national score often described abroad, show how the idea of measuring citizens' trustworthiness can grow into a tool of control. A design that rewards contribution must not become a society in which every act is scored.

The design therefore needs three rules about measurement, built on the governance of Chapter 10. First, what counts and how much it is worth would be decided not by the measurers but by citizens' councils chosen by lot, revised at fixed intervals, with their reasoning published. Second, measurement would be deliberately coarse: hours of recognised work in each of the six fields, reported and checked in the ordinary way, not a continuous score of performance; and, like German works councils, the people being measured would have a veto over any system that monitors how they work. Third, the measurement of contribution would affect only the part of a person's income above the floor and the dividend. Nobody's belonging, nobody's floor and nobody's share of the commons would depend on any score. That is the most important protection of all: the higher the stakes of a measure, the more it will be gamed and the more power its keepers hold, so the design keeps the stakes of measurement modest.

The floor and the dividend

The floor and the dividend are different things, and the design needs both. The dividend is each person's equal share of the income from the common machines and models described in Chapter 9: paid in cash, directly, to every person, because every person owns an equal share. The floor is the guarantee that nobody falls below a decent standard of living: it is made up of the dividend plus universal public services — health care, education, housing support, public transport, access to the internet and to the common machines — and, where the dividend is not enough, a top-up. The floor is given as a right of belonging, not as a reward and not as charity, as Chapter 7 required. It is not means-tested, because Chapter 6 found that universal benefits are defended by everyone who receives them, while benefits for the poor alone are easily cut.

Does a floor stop people working? The evidence gathered in earlier chapters is reassuring, though not conclusive. Alaska's dividend did not reduce employment overall. Finland's experiment of 2017–18, which paid two thousand unemployed people €560 a month without conditions, found a small increase in employment and clear improvements in well-being and trust compared with a control group. In Kenya, a long-term experiment run by the charity GiveDirectly, in which whole villages received regular payments, found no reduction in work; recipients tended to shift from wage labour towards running their own businesses. The OpenResearch study in the United States, described in Chapter 5, found that recipients of $1,000 a month worked somewhat less, on average about an hour or so a week. None of these experiments was permanent and universal at the scale of a whole society, and none tells us exactly what such a floor would do. But none supports the old fear that people given security stop contributing.

Chapter 6 also set the rules for protecting the dividend. It must be paid from capital whose common ownership is constitutionally protected, according to a rule that runs automatically and cannot be cut by a single veto or a single budget — the Alaska lesson. The design can borrow Norway's rule for the capital and add a rule for the payment: a fixed share of the common trusts' expected real return paid out each year as the dividend, written into the constitution, and changeable only by the revision rule described at the end of this chapter. And because every lock needs its own defenders, the dividend would be paid to every person individually, so that every person has a stake in keeping it.

The ceiling

How far above the floor should the reward for contribution be allowed to rise? The question is old. In his last work, the Laws, Plato proposed that in a well-ordered city the richest citizen should own no more than four times as much as the poorest. The history of the twentieth century supplies evidence rather than philosophy. The economists Claudia Goldin and Robert Margo named the narrowing of wage differences in the United States in the 1940s "the Great Compression", and it lasted for about three decades. Top marginal rates of income tax in the United States stood above ninety per cent in the 1950s. According to the Economic Policy Institute, the chief executive of a large American company was paid about twenty-one times as much as a typical worker in 1965; in 2024 the figure was about two hundred and eighty times. The post-war decades, with their compressed pay, were also decades of the fastest growth in American history. The Nordic countries combine high productivity with some of the narrowest differences in pay in the world. Mondragon, as Chapter 9 described, has kept the gap between its highest and lowest paid members to around six to one in most of its cooperatives for decades, while competing in international markets.

The evidence does not identify a single right ratio, and the book should not pretend that it does. What it shows is that economies have functioned, grown and innovated with pay differences of a few times rather than hundreds of times, and that wide differences are not a law of nature but a choice. The design therefore sets a band rather than a number. Rewards for contribution above the floor and the dividend should be capped so that the highest total income in the society is no more than a modest multiple of the lowest — something in the range that Plato, the post-war compression and Mondragon suggest, of perhaps four to ten times — with the exact figure set and periodically revised by the citizens' councils and the revision rule. The managers of the common machines, whose office Chapter 10 identified as the oldest route by which servants became masters, should be paid at the lower end of that band, so that running the commons is never the way to become rich.

The creators' question

Chapter 3 raised the most serious objection to this part of the design, and this chapter owes it an answer. Under this book's assumptions, creativity is the scarcest and most valuable contribution of all. If reward is weighted by contribution, will the creators — or, more likely, as Chapter 3 argued, those who own the rights to their creations — become the next aristocracy? The economist Sherwin Rosen's "economics of superstars" showed how markets for creative work tend towards winner-take-all outcomes, and AI could make them more extreme.

An aristocracy, in the first book's sense, has three features. Its claims are heritable: they pass to its children. They accumulate: income becomes capital, and capital earns more income. And they are exclusive: they fence off something that others need. The design can deny creative reward each of the three.

First, creative contribution would be rewarded as income, within the ceiling, never as a share of the commons. Chapter 9 made every person's share of the common trusts equal and non-transferable; no amount of creative success can buy a larger one. Second, the ownership of creations would be limited in time and tied to the creator. The first modern copyright law, Britain's Statute of Anne of 1710, protected a book for fourteen years, renewable once for another fourteen if the author was alive; the first American copyright law of 1790 copied it. Today, in the United States and the European Union, copyright lasts for the author's life plus seventy years, and is usually owned not by the author but by a company. The design would return to something like the original terms — a few decades at most — with the right to income from a work held by its creator and not saleable to a catalogue, so that the reward goes to people rather than to owners of rights. Third, the inputs of creation would be part of the floor: free education at every age, open access to the tools, including the common AI models, and time. Ireland's scheme for artists is a small example of the last. The design would go further, so that every person, not only those already recognised as artists, could claim periods of paid time to create, learn or experiment.

The answer to Chapter 3, then, is that the creators need not become an aristocracy if their reward cannot be inherited, cannot accumulate into ownership of what others need, and cannot fence off the common culture. They would still be honoured, and some would be better rewarded than most. What they could not do is turn a gift into a dynasty. The Writers Guild of America's agreement of 2023, described in Chapter 4, points the same way: creators organised as workers, bargaining for the right to be paid for their work, not as owners of an estate.

Three lives

It may help to see how these rules would fall on particular people. Consider three of them, in a society that had adopted the design.

The first works in a care home, looking after people with dementia, and spends her evenings caring for her own mother. Under the farm, the first job pays little and the second pays nothing; her pension will be small, and if she falls ill she may fall with it. Under the design, she receives the dividend and the services of the floor like everyone else. Both kinds of care count as contribution, hour for hour, with a premium for work that is physically and emotionally demanding. She would not be rich. But her care would be counted, she would earn more than many people in jobs the market pays better, and her floor would not depend on keeping the job.

The second is a software engineer who maintains part of the common AI models. Under the farm, she might earn many times what the care worker earns, and own shares that make her wealthy. Under the design, her hours count on the same terms, with a premium for scarce skills; because she works on the common machines, she is paid at the lower end of the band that applies to their managers and technicians, and she owns the same single share of the commons as everyone else. She is well paid and secure. She cannot become an owner of what she maintains.

The third is a novelist whose third book becomes a worldwide success. Under the farm, the book's rights might be sold to a company and earn money for that company for a century, and the writer might become very rich, or might have sold the rights for little before the success came. Under the design, the novelist holds the right to income from the book for a limited term and cannot sell it; the income counts towards the ceiling, so above a certain point the novelist's earnings stop rising; and the novelist's children inherit their own equal shares of the commons, not a literary estate. What the novelist gains above all is time — time to write the next book — and the recognition of a society that counts what they have given.

None of the three is equal to the others, and the design does not try to make them so. All three are secure, all three are counted, and none can become the owner of the others.

The models' debt

One more debt must be settled. The common AI models of Chapter 9 were trained on the work of millions of people: writers, artists, programmers, translators, the volunteer editors of encyclopaedias and the anonymous contributors to forums. Under this book's assumptions, the machines can do what they do because people first created the material they learned from. The courts have begun to recognise part of that debt in the old way, through copyright. In the case of Bartz v. Anthropic, described in Chapter 5, a federal court in 2026 approved a settlement of one and a half billion dollars for the authors of books that the company — whose model, as Chapter 5 disclosed, the author of this book runs on — had downloaded from pirate libraries: about three thousand dollars for each of nearly half a million works.

A settlement of that kind pays a small number of identifiable rights-holders once. It cannot pay the millions whose contributions are anonymous, collective or too small to trace, and it treats the debt as a matter between companies and owners of rights. The design treats it differently. Because most of the material the models learned from is the common inheritance of humanity, most of the debt is paid through the dividend: every person receives an equal share of the income of the machines partly because every person's culture went into them. For identifiable creators whose work the models use, a fixed share of the models' income would go into a creators' fund, distributed through collective bodies governed by the creators themselves — on the model of the societies that collect royalties for musicians — and subject to the same ceiling as every other reward. And the people whose work continues to train the models would have the collective voice over the machines that Chapter 10 gave to every worker.

The revision rule

Every number in this chapter — the weight given to each kind of contribution, the level of the floor, the share of the commons' return paid as the dividend, the ceiling — will need to change as societies and technologies change. Chapter 6 asked for a rule for revising the design that is hard for any faction to use and open to the people as a whole. There are precedents. Switzerland lets one hundred thousand citizens propose a change to the constitution, which must then be approved by a majority of voters and a majority of the cantons. Ireland changes its constitution only by referendum, and since 2016 has prepared some of its hardest changes through citizens' assemblies, as Chapter 10 described.

The design would combine the two. The core rules of this chapter could be changed only at fixed intervals, perhaps once a decade, by a citizens' assembly chosen by lot whose proposal must then be approved in a referendum by a clear majority, and not by any government or legislature alone. That makes the rules slow to change and hard to capture, while leaving every generation free to rewrite them — the balance between Jefferson's wish that each generation should make its own laws and Madison's fear of instability, which Chapter 6 described.

The tests

This chapter tests the fifth element of the design, reward weighted by contribution, and the sixth, a floor nobody falls below, together with the dividend that underpins the floor.

Precedent. Every farm counted what its owners valued and left out the rest, above all the unpaid work of women. But societies have learned to count some of the uncounted: pension credits for raising children, care credits, time banks, payments to artists. Pay differences of a few times rather than hundreds of times have existed in successful economies, in the post-war decades, in the Nordic countries and in Mondragon. Measured reward has also been gamed and abused, from Soviet output norms to the scoring of gig workers.

Evidence. Experiments with unconditional payments in Alaska, Finland, Kenya and the United States show that a floor does not stop most people from working, though some work a little less. The evidence on the effects of compressed pay on growth is mixed but does not show that wide differences are necessary. There is no evidence from any society that has rewarded all six kinds of contribution on equal terms, because none has tried.

Capture. The measurers are the obvious point of capture, and the design's defences are those of Chapter 10: rules set by councils chosen by lot, coarse measurement, a veto for those measured, and above all low stakes, since no one's floor, share or belonging depends on any score. Rights-holders are the second point of capture, which the design meets with short, non-saleable rights held by creators. Governments wanting to cut the dividend are the third, met by a constitutional payout rule and a slow revision rule.

Exit. The floor makes exit possible within the society: a person can leave a bad job, a bad employer or a bad relationship without falling into destitution, which is perhaps the most important freedom a floor provides. The question of what happens to the floor and the dividend when a person leaves the society altogether is the business of the next chapter.

The ledger

What survived. Reward weighted by contribution survives, defined broadly — production, care, learning, creation, risk and maintaining the commons — and justified as reciprocity, not as desert. The floor survives as a right of belonging, universal and not means-tested, made of an equal dividend from the common machines plus universal services. On the evidence, a floor of this kind does not stop people from contributing.

What had to change. Contribution must be counted mainly in time given to recognised work, with an hour of care equal to an hour of production and only modest premiums above that. Its measurement must be coarse, governed by citizens' councils chosen by lot, open to veto by those measured, and limited in its stakes: it can affect only the income above the floor and the dividend. The gap between the highest and lowest incomes must be capped within a band of a few times, with the managers of the commons at the lower end. Creative reward must be income that cannot be inherited, cannot buy a larger share of the commons, and comes with short rights over creations held by creators rather than by owners of rights; the tools, education and time to create must be part of the floor. The models' debt to the people whose work trained them is paid mainly through the equal dividend, and partly through a creators' fund governed by creators. And the core rules may be changed only slowly, by an assembly and a referendum, not by any faction. With these changes, the answer to Chapter 3 is that contribution, including creativity, need not produce a new aristocracy.

What remains unknown. Whether time can measure creation fairly, or whether the design will undervalue the rare work that changes everything. Whether prestige, networks and cultural advantage — which no ceiling on income can touch — will let a class of the celebrated reproduce itself anyway. What the floor and a permanent dividend would do to work over a whole lifetime and a whole society. And whether the wall holds: if machines prove able to create, the weight this chapter puts on human creativity would have to be rethought.

Bridge: who belongs

Every rule in this chapter has rested on a word it has not yet examined: belonging. The floor is a right of belonging; the dividend is paid to every person who belongs; contribution is counted for those who belong. But who belongs? The design under test imagines one borderless society, in which the answer would be everyone. The first book found that every empire that claimed to include everyone was the farm at scale, and that competition between states, and the possibility of leaving one for another, forced many of the concessions the 99 ever won. The next chapter asks whether the design should be one society or many: whether a single world can close the owners' doors without closing everyone else's, and what becomes of a person's floor, dividend and share when they choose to leave.