Putting-out system

Why did merchants scatter cloth-making across a hundred cottages instead of gathering it under one roof?

For roughly three centuries before the factory, the standard way to make cloth in Europe was this: a merchant bought raw wool, rode out into the villages, left a portion at each cottage door, came back later for the yarn, carried it to other cottages to be woven, then to a fuller, then a dyer, and finally sold the finished piece in a distant town. He owned the material at every stage. He owned no workshop at all.

To modern eyes this is obviously inefficient — all that riding, all that carting, no supervision, no consistency. So the question is not why the factory eventually replaced it. The question is why anyone with capital would deliberately choose to disperse a process he could have concentrated.

Take the process apart first, because the answer lives in the seams. Making broadcloth in 1700 meant sorting, scouring, carding or combing, spinning, weaving, fulling, napping, shearing, and dyeing. Ask of each stage: what does it actually need? Spinning needs a wheel, a pair of hands, and light. Weaving needs a handloom, a room tall enough for it, and a skilled operator. Fulling needs water power, so it happens at a mill by a stream and always did.

Now notice something about that list. Almost every stage is separable. The output of spinning is yarn, which is countable, weighable, storable, and portable. Nothing about spinning requires it to happen next to weaving in space or time. When a process has clean joints like that, what forces would push it to be concentrated anyway?

Only two, really: shared equipment too costly to duplicate, and coordination that has to happen face to face. In 1700 neither applied. A spinning wheel was cheap enough that peasant households already owned one. And the coordination — who needs how much of what, and by when — was a matter of tallies and dates, which one merchant with a ledger could hold in his head.

So the process refused to be centralised. But that only tells us the factory was unnecessary; it does not tell us why dispersal was positively better. For that, ask where the labour was. It was rural, and it was already occupied — but seasonally. A farming household is flat out at harvest and comparatively idle for months afterwards. What is the wage a household will accept for work in its idle season, on its own hearth, done by women and children between other tasks? Far below what a full-time urban worker must be paid to live, because it is filling time that had no other buyer.

Two further advantages fall out once you look. The rural cottage sat outside urban guild jurisdiction — beyond the reach of guild rules on apprenticeship, on who may weave, on wages and on quality. And the merchant carried no fixed costs at all: no building, no heating, no idle plant. Cloth demand swung violently with harvests, war and fashion. What does a merchant with no factory do when demand halves? He simply puts out less. What does a factory owner do? He keeps paying for a building either way.

Now put those together and the real division of labour comes into view — and it is not the one people usually name. The physical stages stayed dispersed because nothing joined them. What got concentrated was the one thing that genuinely gains from being in a single head: knowing which markets want what, holding the credit that lets raw material sit unsold for a year, and sequencing a dozen independent hands so the yarn arrives when the weaver is free. The merchant specialised in coordination and capital. Everyone else specialised in one physical operation. That is a division of labour every bit as real as Adam Smith's pin factory — it just had a hundred roofs instead of one.

Why did it end, then? Not because merchants finally noticed the inefficiency. It ended when the machinery changed shape. Hargreaves's spinning jenny of the 1760s was small enough to stand in a cottage, and it stayed there — dispersal survived it. Arkwright's water frame of 1769 was not. It needed a river, a building, and continuous power, and that reintroduced the one condition that had been missing: equipment too expensive and too immobile to duplicate per household. Concentration followed the machine, not the other way round.

That last point deserves a proper hearing, because it is the system's genuine weakness. Material advanced to a household could be quietly kept back, stretched, or wetted to make up weight. Work went at the household's pace, not the merchant's, so delivery dates were soft. Quality varied cottage by cottage. These are what economists would now call agency costs, and they were the price paid for cheap, elastic labour.

It is also worth flagging an argument historians have not settled. One account says the factory won because power machinery demanded it. Another, associated with Stephen Marglin, says supervision and control were themselves a large part of the motive — the factory made workers' effort visible and their pace the owner's to set. The two are not exclusive, and the honest position is that both mattered and their relative weight is still disputed.

• Why proto-industrial regions were often the ones with poor soil, and what that says about the wage the system relied on.
• How the same logic reappears in modern outsourcing and gig platforms, and where the analogy holds up.
• What happened to the households when the work left: the rural de-industrialisation that followed the mills.

Every term the collection defines is gathered in the glossary.