Company Towns: Paid in Scrip, Housed by the Boss, Indebted to the Store

It is one thing for an employer to own your working hours. It is another for the same employer to be your landlord, your grocer and your creditor at once. That was the arrangement in the company towns that spread through the mining valleys, textile districts, timber camps and railway yards of the nineteenth and early twentieth centuries — settlements where a single firm owned the mill, the houses, the shop, the school, sometimes the church and the doctor, and where the money paid out on Friday could be designed to come back by Monday. I keep returning to these places because they are the cleanest illustration I know of a principle that has nothing to do with coal: power is not measured by any one relationship you have with someone, but by how many of your relationships run through the same hands.

What a company town actually was

The pattern showed up wherever work had to be done far from an existing town. You cannot dig coal where the coal is not, or saw timber where the forest is not. So the firm brought the town to the work: rows of houses, a store because there was no other store, a school because there was no other school, a doctor because the nearest one was a day away. Read charitably, it was infrastructure. Read plainly, it was also enclosure — every service a worker needed to live was now a line item on one employer’s ledger.

These places were common in the American coalfields and in New England mill villages, in British mining and manufacturing districts, and in plantation and industrial settlements across India. The best-known Indian example runs in a different direction — the steel town at Jamshedpur, laid out from 1908 and renamed after the firm’s founder in 1919. It is routinely held up, with some justice, as a model of private planning: wide streets, water, schools, hospitals, decades before any law required them. That example matters precisely because it complicates the story. Not every company town was a trap.

But every company town had the same structural feature, whether it was run generously or brutally: one owner sat astride the worker’s wage, home, food and credit simultaneously. Generosity, where it existed, was a choice the owner could revise. That is what makes this a story about structure rather than about villains — a theme I take up at more length in how technology gets captured.

The wage that came back: scrip and the company store

The sharpest instrument in the company town was not the rent. It was the money — or rather, the thing standing in for money.

Under what the British called the truck system, employers paid workers wholly or partly in goods, tokens or vouchers rather than coin. In the American coalfields the tokens were known as company scrip: privately issued currency, often stamped metal discs or paper notes, redeemable at the company store and nowhere else. A miner short of cash before payday could draw scrip — an advance against wages he had not yet earned — and spend it at prices the company set, in a shop with no competitor for miles.

Follow the arithmetic and you see why this was so powerful. The wage leaves the company as scrip, is spent at the company store, and returns to the company minus nothing. Rent is deducted before the worker sees anything at all; so are medical fees, tools and fuel. If those deductions exceed the week’s earnings, he does not get a small pay packet — he gets an empty one, and a balance carried forward. During the debates that led to American wage legislation in the 1930s, legislators were pointed at accounts of exactly that: envelopes containing nothing for a full week’s work, because store charges had eaten the lot.

The obvious abuse was price: a store with a captive market can mark up flour and boots as it likes. The subtler abuse was the ledger. A worker who ends the month owing the firm is not simply poor; he is encumbered. He cannot walk to a better employer without settling up, and he cannot settle up on wages that never arrive in a form he can save.

A wage you can only spend in one shop is not really a wage. It is a rebate the employer pays itself, with your labour in between.

Lawmakers understood this early. Britain’s Truck Act of 1831 — consolidating and extending earlier attempts — required that wages in the covered trades be paid in the coin of the realm, and barred employers from dictating where that money was spent. It was a genuine landmark. It was also, for a long stretch, close to a dead letter: enforcement was thin, penalties were light, and a worker who sued his employer could expect to lose his job. In the United States, payment in scrip was finally ruled out of bounds by federal wage law in 1938, which treats scrip, tokens and similar devices as improper mediums of payment; even so, there are accounts of such systems limping on in pockets for years afterwards. The lesson I draw is not that the law failed, but that writing the rule is the easy half. Rules against concentrated dependence only bite when someone can afford to invoke them — which is exactly what the company town was designed to prevent.

Why leaving was so hard

People ask, reasonably, why anyone stayed. The answer is that four separate locks were fitted to the same door.

• Debt. A balance owed at the store follows you. Leaving with it unpaid could mean forfeiting back wages, or worse.
• Tied housing. Quitting the job meant quitting the house, with a family in it. The notice period on a home should never be the same as the notice period on a job; in a company town it was.
• Remoteness. Mines and mills sat where the resource sat. The next employer might be a hundred miles off, and the fare was money you did not hold in spendable form.
• Monopsony. Even if you got out, there was often only one buyer of your skill in the region, and one buyer sets the price.

Any one of these is survivable. Stacked, they convert a contract into a condition. And none of them requires cruelty: a courteous employer, applying ordinary commercial logic to each lock separately, produces a worker who cannot leave.

Pullman, 1894: the cut that was not shared

The flashpoint everyone remembers is Pullman, Illinois — a planned town south of Chicago built by the sleeping-car manufacturer, with brick housing, parks and a library that were, by the standards of the 1880s, unusually good. The firm owned all of it, and rent was deducted from wages before they were paid.

When the depression that began in 1893 hit orders, the company cut wages — accounts differ on the exact severity, with the commonly cited figure around a quarter and some sources describing deeper cuts for particular trades — and did not cut rents. That asymmetry is the whole story in one line. The firm treated its manufacturing arm as a business exposed to the market, and its landlord arm as an investment entitled to its return. The same people absorbed both decisions.

Around four thousand workers walked out in May 1894. The American Railway Union, led by Eugene Debs, backed them by refusing to handle trains carrying the company’s cars, and a local dispute became a national rail stoppage. The federal government intervened with an injunction and troops; the strike was broken and Debs jailed. But Illinois then went after the arrangement itself, arguing that owning and running a town exceeded the company’s charter. The state Supreme Court agreed in 1898 and ordered the firm to sell its non-industrial holdings — though compliance was slow, the sales running to the end of that decade.

I find the court’s reasoning more interesting than the strike’s outcome. It did not rule that the housing was bad, or the rents extortionate, or the owner wicked. It ruled that a manufacturer had no business being a municipality. The objection was to the combination.

Being fair to the paternalists

It would be easy, and wrong, to write every company town up as a scheme. Some firms built housing better than anything workers could otherwise rent, ran schools where the state ran none, piped clean water into places that would not have seen it for another generation. Workers sometimes chose these towns over the alternatives, and said so. The paternalism was frequently sincere.

The problem was never that every owner was cruel. It was that the arrangement made a worker’s entire life contingent on the owner’s continued benevolence, and benevolence is not a right you can enforce. A good landlord who is also your employer can still evict you for arguing about your wage. Concentration is the hazard; conduct is only the weather. The same distinction runs through the Standard Oil monopoly — the case against it was not that its kerosene was poor, but that one firm controlled the pipeline, the refinery and the freight rate together — and through the Triangle Shirtwaist fire, where the locked stairwell was not sadism but an ordinary anti-theft measure taken by people who had no reason to weigh their workers’ escape against their own convenience.

Ask not whether the owner is kind. Ask how many of your exits they control — and what happens to the kindness on the day their margins move.

The modern rhyme

Nobody hands out metal tokens now. The pattern persists anyway, because it was never really about tokens.

Employer-tied housing has not gone away — it is standard in construction camps, domestic work, plantation labour and plenty of migrant employment, where losing the job means losing the bed the same night. Employer-tied immigration status is a stronger lock than any company store ever was: when your right to remain in a country is attached to a single sponsor, the cost of complaining includes deportation, and everyone in the arrangement knows it. Recruitment debt does the work that scrip used to do; a worker who borrowed to get the job is not free to leave it.

Then there is the softer version most readers will recognise. The company phone, the company account through which your contacts, your calendar and a decade of your correspondence flow — revocable at termination, instantly. Health cover attached to employment. Stock vesting on a schedule that makes leaving expensive at every moment except one. None of it is scrip; all of it increases the number of things that end when the job ends.

The sharpest modern echo is the platform that is simultaneously your workplace, your payment rail and your reputation. A driver or courier or seller on such a system does not merely take work from it: their earnings are held and disbursed by it, and their access to future work depends on a rating it computes and owns, which cannot be carried to a competitor because it was never theirs. Deactivation is not a dismissal you can appeal to a manager who knows your name; it is a state change in a database. When one entity sets your pay, holds your money, scores your conduct and can end your livelihood without a conversation, you are in a structure the Pullman workers would have recognised — minus the brick houses. The monitoring layer that makes such scoring possible is its own subject, which I take up in bossware.

The test worth keeping

The useful question about any of these arrangements is not whether the employer means well. Most do, most of the time. The useful question is the one the Illinois court effectively asked in 1898: how many dependencies are stacked on one counterparty, and what happens to me on the day that counterparty’s interests and mine diverge?

If a single decision by a single firm can simultaneously remove my income, my home, my savings, my standing and my ability to be hired elsewhere, then I am not a party to a contract in any meaningful sense. I am a resident of a company town. The scrip is optional. The concentration is the thing.