Every worker who moves into a factory is a worker who has stopped growing food, and somebody has to make up the difference before the move can happen at all.
The previous lesson ended with a number that makes this concrete: by around 1710, on the Cambridge Group's reconstruction of English occupations, only about 35 per cent of the male labour force worked in agriculture. That is a pre-industrial society in which two men in three are already doing something else. This lesson asks how the food got produced and where the extra people came from, because both had to be settled before a single spinning machine mattered.
The problem as a ratio
Put the occupational figure to work. If 35 per cent of workers farm, then each farm worker is producing food for himself and for the 65 per cent who are not, so on average he supports people's worth of food, counting himself. By 1851 the agricultural share was under a quarter, call it 22 per cent, and the same calculation gives 4.55. Output per agricultural worker therefore had to rise by a factor of about 1.6 over that century and a half simply to hold the country's food supply level per person.
Except that it did not have to hold food level per person. The population was more than doubling at the same time, and living standards, in the end, rose. So the true requirement on English agriculture was larger than 1.6, and the gap could be closed in only three ways: grow more per acre, farm more acres, or import.
All three happened, in that order of importance for most of the period, and the third eventually overtook the others in a way that decided British politics for a generation.
What actually changed in the fields
The central technical change is easy to state and easy to underrate: the fallow disappeared.
Under the medieval three-field system, a third of the arable lay idle each year to recover its fertility. That is not superstition. Cereals draw nitrogen out of the soil, there was no way to put nitrogen back, and resting the ground let weeds and grazing animals restore a little of it. The cost is severe, since a third of the best land grows nothing.
The Norfolk four-course rotation removed the fallow entirely by giving each field a useful crop every year: turnips, then barley, then clover, then wheat. Turnips are a root crop that can be hoed clean of weeds and fed to livestock through the winter, which had previously been the season when animals were slaughtered for lack of fodder. Clover is a legume, and legumes host bacteria that fix nitrogen from the air into the soil. Nobody in eighteenth century Norfolk knew that, since the nitrogen cycle was not understood until the 1880s, but they could see that wheat after clover did well.
The livestock half of the loop matters as much as the crops. More winter fodder means more animals kept alive, which means more manure, which means more nitrogen returned to the arable. The rotation is a machine for accumulating fertility, and it works whether or not the farmer can explain it.
Example. Take a hundred acres of arable under a three-field rotation with one field fallow, and convert it to a four-course rotation with no fallow. How much does the cropped area rise, and why is that not the whole gain?
Under three fields with one resting, two thirds of the land carries a crop, so 66.7 acres. Under the four-course all 100 acres do. The cropped area rises by per cent, and that is before any change in yield per cropped acre. It is not the whole gain because two of the four courses feed animals rather than people, so the extra 33 acres are not 33 acres of extra grain, and because the fodder crops raise the number of livestock, which raises manure, which raises the yield on the grain courses in later years. The honest summary is that the direct arithmetic gives a large one-off gain in land use, and the indirect nitrogen effect gives a slower compounding gain in yield, and separating the two in the historical record has proved very hard.
Now you. English wheat yields are usually put at roughly 10 bushels an acre around 1300 and roughly 20 by 1700, before the four-course spread widely. What does that timing imply about the standard story?
Answer
That most of the yield gain was already banked before the famous eighteenth century improvers appear. A doubling between 1300 and 1700 cannot be caused by Townshend, Coke or Bakewell, all of whom belong to the eighteenth century, so the credit belongs to earlier and less celebrated changes: convertible husbandry, marling, floated water meadows, the spread of legumes in ordinary rotations, and steady selection of seed. This is exactly Eric Kerridge's argument in The Agricultural Revolution of 1967, that the revolution happened between 1560 and 1670 and the eighteenth century merely publicised it. The point to carry forward is not that one dating is right but that agricultural improvement in England was long, slow and largely anonymous, and that the named improvers were publicists working inside a process already running.
Enclosure
The change everyone remembers is not a technique but a change in property. Enclosure converted land held in scattered strips in open fields, with common rights of grazing and gleaning attached, into consolidated fields in single ownership with those rights extinguished.
Much enclosure happened privately by agreement over centuries. What is distinctive about the eighteenth and early nineteenth centuries is that it went through Parliament: roughly four thousand private acts, covering something like 6.8 million acres of England, which is about 21 per cent of the country, concentrated between 1750 and 1830. Each act appointed commissioners who surveyed a parish, extinguished the common rights, and reallocated the land in consolidated blocks.
The economic case made at the time was that consolidated holdings could be improved, drained and rotated by an owner who captured the whole benefit, while open fields required collective agreement to change anything. That is a real argument. What is much less clear is whether enclosure delivered the productivity gain claimed for it. Robert Allen's work on the south Midlands found the yield advantage of enclosed over open fields to be modest, and argued that the substantial gains had come earlier from yeoman farmers on open fields, while parliamentary enclosure mainly transferred income from small occupiers to landlords by removing common rights that had real cash value to a cottager.
The cost side is not in doubt. A family with a cow on the common and a right to gather fuel lost both, and the compensation, where any was allotted, was often a plot too small to be worth fencing. Enclosure did not empty the villages by itself, since the enclosed farms still needed labour, but it converted people with a stake in the land into people with only a wage, and that is the population the mills would later hire.
Example. Four thousand acts covering 6.8 million acres, concentrated in the eighty years from 1750 to 1830. What was a typical act doing, and why did the procedure itself hurt the smallest holders?
The averages are acres an act and acts a year. Seventeen hundred acres is about one parish, so the unit of enclosure is the village, decided as a whole and at once, which is why a minority of occupiers in a parish could not opt out. The procedure hurt the smallest holders because it was expensive and the cost was charged to the people who received land. A private bill, a survey, the commissioners' fees and then fencing and ditching the new allotment commonly ran to a pound or two an acre. Set that against a southern agricultural labourer on about ten shillings a week, £26 a year: a cottager allotted two acres faced perhaps £3 of unavoidable fencing, six weeks of wages, for a plot too small to farm. Selling to a neighbour was the rational response, and the concentration of land after enclosure is partly this, a cost threshold rather than a conspiracy.
Now you. Allen found the yield advantage of enclosed over open fields to be modest. Enclosed rents, though, commonly rose steeply, often by half or more. How can both be true, and what does it mean?
Answer
Rent is what the landlord captures, not what the land produces, and the two move apart whenever the change alters who has a claim rather than what is grown. Extinguishing common rights transferred the value of grazing, gleaning and fuel from the commoners to the owner, and that value shows up as higher rent without a single extra bushel of wheat. So a large rise in rent alongside a small rise in yield is precisely what a mainly distributive change looks like, and it explains why landlords promoted enclosure so energetically while the aggregate figures for English agricultural output refuse to show a matching jump. Two honest qualifications. Some of the rent rise reflects real improvement, since consolidated fields were drained and converted to pasture in ways open fields could not be, and separating that part from the transfer is the hard empirical problem in the literature. And the finding is contested: Robert Allen's south Midlands sample is not all of England, and historians who work on other regions report larger productivity effects.
The people
The other half of the problem is where the extra bodies came from, and here the answer overturned a century of assumption.
The obvious guess is that death rates fell: better food, better medicine, fewer plagues. When Edward Wrigley and Roger Schofield reconstructed English population from 404 parish registers, published in 1981, they found the opposite emphasis. Most of the acceleration in English population growth between the late seventeenth century and the early nineteenth came from rising fertility, not falling mortality.
The mechanism is marriage. In an economy with the European Marriage Pattern described in the first lesson, almost all births happen inside marriage and couples marry only when they can support a household. If wages and opportunities improve, people marry younger and fewer stay single, and both effects raise the birth rate without anyone deciding to have more children per marriage. The mean age at first marriage for English women fell from around 26 in the late seventeenth century to around 23 by the early nineteenth, and the proportion of women never marrying fell sharply.
Example. Show roughly how much a three year fall in the age at first marriage adds to completed family size, and what that does to the growth rate.
Marital fertility in this population runs at something like 0.3 births per year of marriage in a woman's twenties. Three extra years of marriage therefore adds about births per woman, close to one extra child per family, on a completed family size of four or five. Feed that through the vital rates: an English crude birth rate rising from about 30 per thousand to about 40 per thousand, against a crude death rate near 28, turns natural increase from 0.2 per cent a year into 1.2 per cent. A change in the average wedding date, spread across a whole population, is enough to take a stagnant population to one that doubles in sixty years, with no change in medicine or in anyone's intentions.
Now you. Why does this finding make population growth a consequence of the economic change rather than an independent cause of it?
Answer
Because the marriage age responds to economic conditions. If people marry when they can afford a household, then rising employment and rising wages pull the wedding date forward, which raises the birth rate, which raises population. The causal arrow runs from the economy to demography, not the other way, which rules out the tempting story in which a mysterious population surge created the demand and the labour that industry then used. Two honest qualifications belong with it. First, mortality did improve somewhat, particularly the retreat of plague after 1665 and of smallpox after inoculation spread, so fertility carries most of the explanation rather than all of it. Second, the mechanism is a general feature of the Malthusian preventive check working normally, so it explains why the population grew but not why the wage failed to fall back, which is the thing still to be explained.
What English farming could not do
Agriculture bought the time, and it ran out of room. England had been a net exporter of grain in the first half of the eighteenth century, encouraged by an export bounty, and became a net importer from the 1760s as the population climbed. Imports were small at first, a few per cent of consumption, and rose steadily.
That turned food into the central political question of the period. The Corn Laws of 1815 blocked imported wheat until the domestic price reached a high threshold, protecting landlords' rents at the cost of the price of bread, and the fight over them ran until repeal in 1846. The economics were plain on both sides: an industrial workforce eats imported food, and the people who own the land do not want it imported. Which side won tells you which interest had become the more powerful, and the answer arrived in 1846.
By the 1870s Britain was importing roughly half its wheat, which settles the question of whether British agriculture fed British industrialisation. It fed the first century of it and then handed the job to the world market, which the country could reach because it had ships, an empire and, by then, the manufactured goods to trade for grain.
What this leaves
England entered the eighteenth century with a farming sector productive enough to release two thirds of its men to other work, a property system that concentrated land and cut cottagers loose from it, and a marriage pattern that turned improving conditions directly into more people.
None of that explains a single machine. A country can have all of it and remain a prosperous agrarian economy, which is roughly what the Dutch Republic was. The preconditions are necessary and they are not sufficient, and the next lesson takes the four serious explanations of why the machines appeared in this country rather than another and tries to break them against each other.