When shirts cost $3,500


That $3,500 figure doesn’t even include the cost of the spinning wheels, buttons, and thread nor the profit margin.

1 Like

Um, okay, point taken, but how many shirts are made today by people making anywhere near $7.25/hour? Aren’t most of the clothes that most Westerners wear produced in what are pretty close to sweatshop conditions?


200 years ago, you’re already in the beginnings of the Industrial Revolution, and the British press was complaining that shop clerks and farm girls could wear the same fashions as the higher end ladies (because of the cheap price of cotton cloth). I would push that claim back at least 50 years…maybe a hundred years, because you can sure as everything spin more than 4 yards a hour with a spinning wheel, which were more in use in the 1700s. Funny you should run this right as I’m in the middle of a weaving project. LOL!


Women & children making clothing in the house weren’t making any money then, either.


Almost certainly, but that is also due to increasing automation.

When making shirts and weaving were skilled jobs requiring long training, the artisan couldn’t be replaced easily and had more control over the profit they took on their labour, as they were the means of production.

Now the machine does most of the work, the human in the loop is less skilled and easily replaced. So that transfers power to those who own the means of production as embodied in the computer controlled machinery.


Ok, so what if the person were making $1/hour instead. That’s still ~$500 for a shirt.


Seems like some of the estimation came from a “dark ages recreation site” so pay it back even farther. 1500 years ago?

That’s still a 20th century wage.

1 Like

Sure, an “artisanal” shirt would today cost $3,500, and I would not be surprised at that price for a completely hand-made shirt. But really, the cost of labor was far lower a thousand years ago, in today’s dollars, than $7.50/hr because (a) human life wasn’t as valued, and (b) there really wasn’t much else a human could do but work for virtually nothing. It’s not like a peasant could aspire to become a lord.


Makes you truly appreciate the machines.


We will only truly appreciate them when access to them and their output is more fairly shared and they don’t threaten our incomes. Currently the machines are helping to make the distribution of wealth even more unequal.

I’m in the interesting/worrying position of watching the slow encroachment of algorithms in my own area of expertise. At the moment I’m like the Medieval artisan, providing an individualised and bespoke service, but can see that aspects of what I do can be automated. I doubt I’ll be completely put out of work in the remaining 20-25 years of my working life, but may no longer be providing individual service by the end of it.


Yes. And the original Luddites were against machines not because they hated machines for their own sake (or feared a gear-driven Skynet) but because their livelihoods as skilled workers were threatened.


However, the number of hours it takes to teach calculus, play a symphony, or palpate a swollen gland has remained largely constant over that period. That means that the cost of health care and education will rise whenever the cost of manufacturing falls – not because health care is getting more expensive, but because everything else is getting cheaper.

Once upon a time, no-one could teach you calculus, because no-one knew it. Then, you could learn it only from the few people who knew Liebnitz and Newton or had read and understood their work. For a long time, you could only learn it at expensive universities. When I were a lad, we got taught it for free. Nowadays you can get Wolfram Alpha to do it for you, without even needing to stay awake in class. I don’t think that’s “remarkably constant”.

Even revealing the hidden costs of ‘free’ education, it should be obvious that teaching & learning is an economic multiplier. One day, only one person knows how to do something; the next day, two do; the next day, four; and so on.

Also, in theory, as automation frees up labour, there are more people available to learn expensive skills, which should make them more common and therefore cheaper. That this happens very slowly, if at all, is sign that there’s someone, somewhere, gaming the system. Though that’s not news.


Everybody on every site I’ve seen run this is fussing over how strictly accurate the $3500 figure is. That’s not the point. It’s only there to provide an easily-graspable weight to “480 hours of human labor,” to help you understand how different cultural expectations were for goods. Nobody’s claiming to nail down a precise exchange rate between 2015 US dollars and [arbitrary Middle Ages time/place] person-hours.


There’s an argument that goldbugs make–an ounce of gold buys a good man’s suit. It’s the kind of argument that seems calculated to interest a particular class (and gender) of person.

1 Like

The thing took 480 hours to make. However, compare it with how many hours of wearing it lasted. Then recalculate the hours-to-make vs hours-to-wear for contemporary garb. Then, and only then, we get data that say something useful.

Also, back then, they didn’t have the “whiter than white” hysteria and “active oxygen” detergents that eat fibers for lunch.


Maybe the other class/gender groups don’t have enough prospective customers to be worth targeting? Marketing 101.

I found this almost painful to read, for the reason that it projects back present day pay (and the productivity of a hobbyist) back to a time when none of that true.

It is naeive about the realities of miserable labour conditions in the textile industry throughout history that continue to this day.

How much were people paid to grow and collect cotton? Nothing, actually. In the U.S., they weren’t paid anything, because for decades, they were slaves.

Was automation the key to cheap cotton mills? Not entirely. In 1819, British mill owners opposed new child labour laws that would ban hiring children under the age of nine, and impose a 12-hour workday. The new laws were rarely enforced.

Children regularly lost limbs and were even decapitated by machinery, and they often died of work-related disease by the age of 25.

In 1911 in the US, the Triangle Shirtwaist Factory fire was the worst industrial disaster in US history, killing 146 people.

The cotton used in a lot of clothing today is still grown in the U.S. - due to subsidies, but lots of that cheap clothing is sewn in sweatshops, places with appalling working conditions, no labour or safety regulations.

The people working in the factory in Bangladesh that killed over 1100 people and injured 2500 more were making $34 a month. That is 20 cents an hour - last year.

After the accident, that has now doubled to $68 a month.

This is a reality of automation not just in textiles, but in plenty of other industries. It’s not automation that makes it cheap, it’s exploitation of workers, because people are considered cheaper and more disposable than machines, because machines cost money.

But as long as we get cheap shirts, right?


That thread count, spinning times, etc all strike me as cotton. Medieval Europe didn’t give two shits for cotton until it’s late period and then it was still not cultivated in Europe. Peasants were more likely rocking wool with lower thread counts, huge stitches like 6-10mm and other compromises, though IMO wool over cotton was not considered a compromise then, because wtf is cotton? That crazy shit from India?

Linen, wool, leather, all easier than cotton. So what if nobility or merchant classes rocked a shirt that cost a month’s worth of a middle-class wage that didn’t exist yet? That does not mean the peasants worked 1-2 months a year simply to continually have 3-4 articles of clothing. It meant they wore the simpler garb and if they were lucky, worked hard and tended their clothes well, they had a nicer article aside.

These people are at a wedding, they got their best on, and it’s probably all wool & linen, not one stitch of cotton.

That’s not to diminish the incredible differences btwn pre & post industrial, just that this is a bit fudged for shock value.