1. The New World the Spinning Frame Made
In 1769, Richard Arkwright obtained a patent for the water frame.
In 1771, the world's first water-powered cotton-spinning mill rose at Cromford in Derbyshire. A five-storey stone building. The waters of the River Derwent turned a waterwheel, and the waterwheel drove the spinning frames. When you stepped inside, the sound came first — the mechanical roar of hundreds of spindles rotating at once, rising through the stone floor and into the soles of your feet. Then the smell: machine oil, cotton-fibre dust, sweat — all mingled in the air. Windows were rarely opened. Cotton thread could be spun without snapping only in hot, humid conditions. In summer the mill was a furnace; in winter it was still far hotter than the air outside. There was no way to keep the cotton fibres in the air out of the lungs.
Roughly three hundred people worked there. Most were women and children. As of 1771, children as young as seven worked thirteen hours a day, six days a week. By 1800, when the workforce had grown to 1,150, some 700 of them — sixty-one per cent — were classified as children.1
The task given to the youngest was usually that of the "piecer." They moved between the spinning frames, tying together broken threads. The work required quick fingers and small bodies. Because the thread had to be joined without stopping the machine, children reached their hands between moving spindles. Lateness was punished with fines. Falling asleep, with beatings. Children who dozed off beside a spinning frame had their hands caught in the machinery.
Arkwright's mill was a prototype of a new world. The model was replicated across Lancashire, into Yorkshire, into Manchester. By the early 1800s, thousands of mills had been built throughout Britain.
Book 1 examined this transformation from a different angle. The handloom weavers of Yorkshire losing their livelihoods to the spinning frame — an estimated tens of thousands of handloom weavers across England and Wales (cotton, wool, and linen combined) in the 1780s,2 gradually undercut by machine prices, falling below subsistence wages by the 1830s, and virtually extinct by the 1850s. Arkwright building a new order of production. Technology redefining the meaning of labour.
Book 4 looks at the same era through the lens of institutions. Technology arrived in 1769. When did institutions follow?
2. Law on Paper (1802–1833)
Thirty-three years after the mills began to spread, in 1802, Britain's first piece of factory legislation appeared.
The Health and Morals of Apprentices Act. The first irony was that the man who championed the law was himself a mill owner, Robert Peel Sr. An epidemic had broken out in his factory. When the risk arose that sick children might spread disease to local residents, he requested regulation himself. It was a business decision.
The telling detail is the law's scope. The 1802 Act applied only to "parish apprentices" in cotton and woollen mills. Children sent from the workhouse to the factory — children with no guardian, children who could not escape their contracts — were the only ones covered. "Free-hire" children who lived with their parents were not covered at all.3
The provisions sounded reasonable: a twelve-hour ceiling on apprentice labour, mandatory ventilation, mandatory schooling, separate sleeping quarters for boys and girls.
One thing was missing. No one to enforce it.
The law gave supervisory authority to local magistrates. But the magistrates came from the mill-owning class. The regulated became the regulators. Just as the Roman Senate had neutralised Augustus's enforcement of law, English mill owners sat on the magistrates' bench and oversaw their own factories.
The 1802 Act existed for thirty-one years. Yet over those thirty-one years there is almost no record of its ever being enforced.4
In 1819, a second law was passed. The Cotton Mills and Factories Act. Employment of children under nine was prohibited. Once again, there was no enforcement mechanism. The law existed. Inspectors did not.5
A law without an enforcement mechanism is words on paper. The Acts of 1802 and 1819 proved as much. Thirty-one years of failure gave birth to the central innovation of 1833.
3. The New Lanark Experiment
In 1800, Robert Owen became managing partner of the New Lanark mills in Scotland.
A record survives from the day he first arrived. The mill village sat wedged in the gorge of the River Clyde. Four spinning mills towered above the sound of the water. Two thousand workers had built their homes around the mills. When Owen walked inside, what he found were children — five, six years old. They stood working thirteen hours a day. Piecing threads, crawling under spinning frames to clear cotton dust, shaping their bodies to the rhythm of the machines.
Owen began making voluntary changes. He raised the minimum working age to ten. He reduced hours to twelve — pushing back, little by little, against his partners' resistance. In 1816 he established Britain's first infant school. There was no corporal punishment. There was music and dancing. The children learned about nature and history. The walls of the schoolhouse were painted with pictures of animals. In 1817, he offered a slogan: "Eight hours labour, eight hours recreation, eight hours rest." It is hard to imagine how radical that sentence was in early-nineteenth-century Britain, when the average working day was fourteen to sixteen hours.
The experiment proved two things at once.
First, child welfare and profitability were compatible. Owen's New Lanark turned a profit. Dividends were paid to the partners, and visitors poured in from across Europe. Royalty and aristocrats travelled down to a Scottish gorge to see the experiment for themselves.
Second, voluntary reform does not spread. Competing mills did not follow suit. Factories that exploited children saved on costs and held the advantage in the market. A single good-hearted manager could not change the structure. Without legal compulsion, good practice loses to bad practice.
Owen submitted a draft bill directly to Parliament. No employment under ten, a six-hour limit for children under twelve, compulsory schooling. His proposal was delayed until 1819, and the law that finally passed was a greatly weakened version. What emerged was legislation that fell short of what the voluntary reformer was already doing.
4. Sadler and the 682 Pages
In 1831, the Member of Parliament Michael Sadler proposed a select committee to investigate the conditions of child labour. The hearings began the following year.
Eighty-seven witnesses testified. The result was a record of 682 pages.6
Elizabeth Bentley, aged twenty-three:
At what age did you begin to work at a factory?
"Six years old."
What were the hours of labour during busy times?
"From five in the morning till nine at night."
Was there time for meals?
"Forty minutes at noon."
Time for breakfast?
"No, we ate while we worked, if we could."
How did your mother tell the time?
"She woke between two and four in the morning and asked the miners what time it was."
What happened if you worked too slowly?
"I was strapped."
Peter Smart:
"I was locked in the factory day and night."
"I was bound for six years, and my mother received fifteen shillings."
"I tried to run away and was caught and whipped."
After becoming an overseer: "I was forced to beat the children. They slept only four or five hours."
John Hall, who had observed factories firsthand, testified that children's bodies were being reshaped to fit the machines. He was not himself a child labourer but an observer of factory conditions. The knees of children who crawled under spinning frames to tie threads bent inward; the shoulders of children who stood all day in the same posture rose on one side. "You can pick out the factory children on the street at a glance. Their gait is different." The machine was remaking the human body — a physical precursor to the way algorithms, two centuries later, would reshape human labour and decision-making.
The Sadler Committee's methodology drew criticism. The questions were leading; the selection of witnesses was biased. These were fair objections. Yet the criticism itself revealed something important. Parliament demanded more rigorous evidence, and that demand led to the more systematic field investigation of the Royal Commission (1833). Led by Edwin Chadwick, the Royal Commission combined sworn testimony, direct interviews with children, and medical examinations — and reached the same conclusions. Criticism of method had produced better evidence-gathering. The criticism, too, was part of the system.
Sadler himself lost his seat in the general election of 1832. But the 682 pages remained. They were printed and distributed. Public opinion shifted.
5. The Innovation of 1833
The Factory Act of 1833 was passed.
Its substance was not dramatically different from earlier laws. No employment under nine. An eight-hour ceiling for children aged nine to thirteen. A twelve-hour ceiling for those aged thirteen to seventeen. Two hours of daily schooling for children under thirteen.
But one clause was different.
Four Factory Inspectors and eight Sub-Inspectors shall be appointed.
That was all. What had been born was not a legal provision but an enforcement mechanism. The difference between 1802 and 1833 was not the content of the law. It was the creation of people whose job was to enforce it.
Four inspectors and eight sub-inspectors. Twelve people to cover every factory in Britain. A physically impossible scale. Yet the mere fact of their existence served as a deterrent. Not knowing when someone might arrive was enough to change behaviour.
In Chapter 8, we encounter the same question two hundred years later. Who audits the algorithms? Does an inspector exist who can verify how COMPAS's 137 variables actually work? In 1833, the answer was twelve people. In the 2020s, the answer does not yet exist.
6. Leonard Horner
Leonard Horner was the most influential of the first four Factory Inspectors.
His jurisdiction was all of Lancashire. Manchester, Burnley, Oldham, Rochdale, Ashton. He was perpetually on the move — by stagecoach, and on rainy days on foot through mud. From the moment he knocked on a factory door, resistance began.
At some mills, word of his arrival raced ahead of him. He watched children being hustled out through back doors — small, oil-stained hands gripping the doorframe and then releasing, bare feet slapping across wet flagstones and vanishing into alleys. The spinning frames were still turning. The workshop floor was empty. The mill owner stood there, unfazed. No violation here.
At other mills, owners were forging birth records. At others, children had been coached to recite a rehearsed age. Twelve years old. Horner began bringing doctors along. He would take a child's wrist, examine the teeth, check the growth plates, and try to determine actual age — medical age assessment, improvised on the spot. The mill owners fought back in court — did the inspector even have the authority to do this?
His half-yearly reports went up to Parliament. One sentence from those reports testifies to the era: "The devices to which the mill occupiers resort to evade the letter of the law grow more ingenious with each passing season."7
In the early days, Horner tended to defer to the expertise of the larger employers. One had to understand how a mill operated in order to inspect one. But the field changed him. Watching children flee through back doors, confirming forged birth records, standing beside a doctor as he noted the welts from a leather strap. Public protest pushed him further — the Ten Hours Movement led by Richard Oastler and John Fielden was convulsing the factory districts, and workers sent petitions demanding that the inspector enforce the law more firmly. Horner grew increasingly independent in his stance, and at last came into direct confrontation with the mill owners.
The mill owners counterattacked through Parliament. Amendments to curtail Horner's authority were introduced repeatedly. Some passed. Each time, Horner noted it in his reports — regulation was losing its enforcement power. Those records, in turn, became the basis for the next round of reform.
For twenty-six years, from 1833 to 1859, Horner inspected factories.8 His half-yearly reports went up to Parliament. Numbers, names, and factory addresses were recorded. His reports became the evidence base for new legislation. The Factory Acts were amended and expanded in 1847, 1867, and 1878 — from cotton mills to all manufacturing, from children to women and young people, broadening step by step.
Leonard Horner is not famous. He does not appear often in the history books. But he was the man who turned abstract law into actual enforcement. Karl Marx, in Capital, called him "a man who rendered undying service to the English working class."9 Few civil servants of any era have received such a tribute.
The history of the Factory Acts is a history of Parliament, but it is also a history of people like Horner. People who made the law work. Not the clauses on paper, but the people who knelt on the factory floor to verify a child's age. Without them, the law is paper.
7. What Sixty-Four Years Means
From the spinning frame (1769) to effective regulation (1833): sixty-four years.
Compared with Rome's 130 years, that is half. But over those sixty-four years, tens of thousands of children worked sixteen-hour days. The cost of slowness was real.
So why did sixty-four years "work," while Rome's 130 years destroyed the Republic?
The difference came down to three things.
First, the formalisation of evidence. The Sadler Committee's 682 pages. Knowing about a problem and having the problem entered into the public record are different things. Once it is on the record, it becomes harder to ignore. In Rome, the suffering of small farmers never entered the Senate's official records. The visibility of harm determined the speed of legislation — it was no coincidence that the Factory Act (1833) passed the very year after the Sadler Committee testimony (1832) made that harm publicly visible.
Second, social learning. When the 1802 Act failed, the British Parliament analysed why. The conclusion — "there was no enforcement mechanism" — gave birth to the central innovation of 1833: the inspectorate. The system learned from failure. Rome did not learn from the defeat of the Gracchi. What came next was not better reform but civil war.
Third, the design of an enforcement mechanism. Four inspectors like Horner and eight sub-inspectors. A physically impossible scale, yet their very existence changed the structure. The fact that the law could be enforced had the same effect as enforcement itself.10
These three factors turned sixty-four years of slowness into "productive slowness."
The history of the British Factory Acts proves that institutions, even slow ones, can ultimately work. At the same time, it records that the price of that slowness was tens of thousands of children.
8. The Compression of the Speed Gap
Rome's gap was 130 years. Britain's was sixty-four. The gap in twentieth-century financial regulation was thirty-seven years (1973–2010).
History is compressing.
In November 2022, ChatGPT was released. Eight and a half months later, China announced generative-AI regulations. Thirty-three months later, the EU AI Act began its phased entry into force. As of 2026, the United States has no federal AI-regulation law.
The absolute duration of the gap has shrunk. But the speed at which technology changes has accelerated along with it. ChatGPT (built on GPT-3.5) launched in late 2022; GPT-4 followed four months later; within a year after that, Claude 3, Gemini 1.5, and Llama 3 appeared in rapid succession. While regulators deliberated over legislation, the thing they were trying to regulate had already gone through multiple generations. The spinning frames that Horner inspected did not change their basic operating principle in twenty-six years. Today's AI models turn over faster than a bill can pass through committee.
The social learning that took the British Parliament sixty-four years — the failure of 1802, the partial improvement of 1819, the innovation of 1833, the expansion of 1847 — AI regulation must compress into a few years.
We do not have sixty-four years.
In Chapter 3, we turn to a case in which the speed gap was compressed dramatically. From the Black-Scholes model of 1973 to the financial crisis of 2008 — thirty-seven years between financial innovation and financial regulation.
The central figure of that story is not a member of Parliament. It is one person who read a prospectus — alone, from beginning to end.