Okay. I'll go first, because I have a genuine reaction and it isn't a nice one. My first thought reading this was: we are about to spend twenty-five minutes being smug about the Victorians, and we haven't earned it.
That's a strong opening.
It's an honest one. Here's what Daniel wrote in this week. He points out that Jerusalem is a city of cranes right now, high-rises going up everywhere, the light rail chewing through half the city, and he wants to go back to the history books. Specifically, the New York City Subway, the London Underground, and the Paris Metro. Not the engineering, or not mainly the engineering. He wants the daily life of the workers who built them. Which of those labour practices would be illegal and indefensible today. How those same practices contributed to the pace of progress. And then the question that actually bites: has modern construction's greater operational efficiency offset the losses from discontinuing those unsustainable and unsafe practices?
That last one is the whole episode.
It is. And he frames it with a line I want to read properly, because it's the funniest thing in the prompt. He says it might be hard to envision a less coordinated process than the one the City of Jerusalem applies today, but the chaos surrounding the construction of the undergrounds in London and New York was almost absurd by modern standards.
He's not wrong, and I say that as someone who has watched a utility crew stand around a hole on King George Street for a week.
So the comparison set is London, New York, Paris, and the benchmark is the city we're recording in.
Which is the right benchmark. Because the temptation with this topic is to treat it as archaeology. Victorian men with shovels, how quaint, aren't we lucky. That's not what Daniel's asking.
No. He's asking whether we bought something real with the safety, or whether we just got slower and called it progress. Those are different claims and only one of them is true.
Let's start with London, 1863. The Metropolitan Railway. Ten January, eighteen sixty-three. Three and three-quarter miles, Paddington to Farringdon Street, seven stations.
The first one. No playbook.
No playbook at all. And I want to be clear about what that means physically, because I think people picture a tunnel. It mostly wasn't a tunnel. It was a trench dug down the middle of the New Road, which is now Euston Road and Marylebone Road. Thirty-three and a half feet wide. Brick retaining walls on both sides, and an elliptical brick arch spanning twenty-eight and a half feet over the top of it.
So they opened a street like a sardine can, built a brick tube in the bottom, and put the road back on top.
That's cut-and-cover, yes. East of King's Cross there's a seven-hundred-and-twenty-eight-yard tunnel under Mount Pleasant, and then it follows the culverted River Fleet beside Farringdon Road in open cutting. Which matters later.
The Fleet matters later, yes. Who was behind it?
Promoted by Charles Pearson, who was Solicitor to the City of London. He pushed this thing for years. He died in September eighteen sixty-two.
Four months before it opened.
Four months. He never rode it. The engineer was John Fowler.
And the actual digging?
By hand. Navvies, mostly. There was a primitive earth-moving conveyor used to get spoil out, and that's the only mechanisation on the job. Everything else is men with picks and shovels in a trench in the middle of a city.
In a trench in the middle of a city, with the city still running.
Which is where the chaos comes in. May eighteen sixty: a Great Northern Railway train overshot the platform at King's Cross and fell into the workings.
A train fell into the hole.
Same year, a boiler explosion on an engine pulling the contractor's wagons killed the driver and his assistant. May eighteen sixty-one, the excavation collapsed at Euston and did considerable damage to the neighbouring buildings.
Which buildings had people in them.
Presumably. And then June eighteen sixty-two, the one I'd hang the segment on. The Fleet sewer burst after heavy rain and flooded the excavations. Delayed the work by months.
The river they'd culverted, that they were digging alongside, came back.
It came back. That's the thing about this era. Every one of these systems was built on top of infrastructure that already existed and that nobody fully understood anymore. The Fleet had been covered over for decades. They knew roughly where it was. Roughly was not enough.
Thousands of homes destroyed, I saw.
Thousands. Compensation required. And serious concern about undermining and vibration causing subsidence in the buildings next to the line. That's not a side effect. That's the business model. You cannot cut a trench down a road in a dense city without damaging what's beside it.
What did it cost?
Estimated at one million pounds. Actually cost one point three million.
Thirty percent over.
Thirty percent over, in eighteen sixty-three, on the simplest version of the project anyone would ever build.
And it worked.
It worked spectacularly. Thirty-eight thousand passengers on opening day. Nine and a half million in the first twelve months, twelve million in the second. And from May eighteen sixty-four they ran what they called workmen's returns, on the five-thirty and five-forty morning services out of Paddington, at the price of a single ticket. Three pence.
So within a year and a half they'd figured out the working-class commuter.
They'd figured out that the people who needed the railway most couldn't afford the normal fare at the hour they needed to travel. That's a real piece of social design, and it's easy to miss when you're busy being appalled by everything else.
Fine. Now the part I actually want to talk about, because it's the piece I think is the most damning and it isn't a collapse or a boiler. It's the air.
The air is the design failure.
They built it with almost no ventilation.
They built it believing it would use smokeless locomotives. That was the assumption. It didn't hold. They used steam locomotives with condensing equipment, which limited how much steam escaped in the tunnels, but the stations and carriages were notorious for smoke.
Notorious how?
By the eighteen nineties there's a Board of Trade report noting a pharmacist who was treating people in distress after travelling on the railway. He had a remedy for it. He called it the Metropolitan Mixture.
There was a branded medicine for the experience of using the London Underground. That's not a transport network. That's a respiratory event with a timetable.
And the fix is almost as good as the problem. They improved ventilation by cutting openings in the tunnels and taking the glazing out of the station roofs. And the local authorities objected.
To ventilation.
On the grounds that the openings would frighten horses and reduce property values.
They'd rather have the smoke.
They'd rather have the smoke than the hole in the roof, yes. Which tells you something about how the costs of this thing were distributed. The passenger breathes it. The property owner doesn't want to look at the solution.
And the guards.
On the Inner Circle service, from eighteen eighty-four, guards were permitted no relief breaks during their shift. None. That lasted until September eighteen eighty-five, when they were finally allowed three twenty-minute breaks.
It took a year and a half of organised complaint to get a man three breaks in a shift.
On a service that ran in a smoke-filled tunnel.
Right. So that's London. The pioneer. And the thing I want to flag before we move on is that none of this was incompetence. Fowler was a serious engineer. Pearson spent years getting this authorised. They built the thing and it carried twelve million people in its second year. The chaos wasn't a failure of skill. It was the price of doing something nobody had done, in a city that was already finished.
There's a line in Daniel's prompt I keep coming back to, about coordination. London in eighteen sixty is the zero point. There is no coordination to speak of because there's no precedent for what coordination would even mean.
Paris, then.
Paris is the interesting one, because Paris is the opposite. Paris is what happens when you design for speed deliberately, from the first drawing.
Nineteen hundred.
First line, Porte Maillot to Porte de Vincennes, inaugurated nineteenth of July nineteen hundred. Timed for the World's Fair. Construction began November eighteen ninety-eight. And it came out of a project by Fulgence Bienvenüe, dated April eighteen ninety-six.
So four years from approval to running trains.
Under four years, for the first line, in a capital city.
How?
Two decisions. The first is that they rejected the shield method in favour of cut-and-cover, and they said so explicitly. The reason given was in order to speed up work.
They chose the messier method on purpose because it was faster.
They chose the method that tears up the street because the street can be put back quickly. And that's the second decision. Bienvenüe designed a procedure specifically to allow swift repaving of the roads. So the disruption is real but it's short. You're not tunnelling under the street for years with a shaft at one end. You're opening the street, building, closing it, and moving on.
That's a different theory of the project. London was trying to build a railway. Paris was trying to build a railway without losing the street.
And it worked. The tunnels are near the surface, mostly following the twists and turns of the streets above, because deep digging was complicated by the variable ground. Minimum curve radius of seventy-five metres, imposed in nineteen hundred. Which is tight, and even that wasn't held to at Bastille and Notre-Dame-de-Lorette.
Seventy-five metres is tight for a railway.
It's tight. It's the kind of number you pick when the street layout is the constraint and the train is the thing that has to give.
Now the politics, because this is my favourite part of the Paris story.
Go on.
They built the tunnels deliberately narrow. Two point four metre carriages. Specifically so that mainline trains couldn't run on the network.
That's right. It's a defensive move against the state taking the thing over. Paris and the French state had poor relations, historically, and the city wanted a metro it controlled. So they made it physically incompatible with the national network.
They built a railway the wrong size on purpose so the government couldn't steal it.
They also forbade lines out to the inner suburbs, and they made metro trains run on the right, which is opposite to the suburban lines. Every one of those is a guarantee of independence.
That's not engineering. That's a border.
It's a border. And it's worth saying that Paris was designed from the outset as a unified system. Nine lines, later ten. London grew by accretion, companies merging and fighting. Paris was planned.
And the pace held.
By nineteen twenty, despite a few changes from schedule, most of the lines had been completed. There was a competing private network, the Nord-Sud company, concession granted January nineteen hundred and four, built Lines A and B, opened nineteen ten and nineteen eleven. Went bankrupt. Absorbed by the CMP by the end of nineteen thirty.
So the unified plan won.
The unified plan won.
And the disaster?
The Couronnes disaster. Tenth of August, nineteen hundred and three. A fire caused by a short circuit. Eighty-four people dead.
That's the one that changed the trains.
That's the one that made steel carriages mandatory. The carriages were wooden. The fire spread through the train and then through the station. Eighty-four people.
So Paris gets its speed, and it gets a design that was coherent and fast and cheap to build, and three years in it kills eighty-four people in a fire and the regulations change overnight.
Which is the same shape as London's air problem. A design decision made for good reasons at the time, a cost that shows up later, and a regulatory response that only comes after people die.
And it's not only workers. That's the piece I want to hold onto. We've been talking about labour practices, but the passengers were the ones breathing the Metropolitan Mixture and burning in the Couronnes fire.
Both systems. Yes.
New York.
New York is the one where the human cost is sharpest, and it's sharpest because of one specific technique. First underground line opened twenty-seventh of October, nineteen hundred and four, City Hall to a hundred and forty-fifth Street. Ground broken at City Hall on the twenty-fourth of March nineteen hundred, and they called it Tunnel Day.
They had a ceremony for starting to dig.
Contract One, executed February nineteen hundred, between the Board of Rapid Transit Railroad Commissioners and the Rapid Transit Construction Company. John B. McDonald, funded by August Belmont.
And the structure of the job?
Divided into fifteen sections, and subcontractors bid on them. So you've got a fragmented, multi-contractor model. Fifteen separate firms, roughly, working on one line.
That's the coordination problem, right there. That's not a city building a subway. That's a city refereeing fifteen building sites.
And the workers. The sandhogs.
Say what that means, because it's a great word and people use it loosely.
It's the slang for New York's urban miners and tunnel workers. Traditionally Irish or Irish American, and the trade passed down through families. They started with the Brooklyn Bridge in eighteen seventy-two and went on to build the subway tunnels, the sewers, the water tunnels, the Lincoln, the Holland, the Queens-Midtown, the Brooklyn-Battery.
Generations of the same families.
Generations. There's a wonderful detail that Warren Beatty worked briefly as a sandhog on the third tube of the Lincoln Tunnel in the nineteen fifties. He's the exception that proves it's a closed trade.
And the hazard.
Caisson disease. The bends. Work in pneumatic caissons, which means pressurised chambers underwater, exposed workers to decompression sickness when they came out too fast.
Explain the caisson, because I think people hear pressurised chamber and picture a submarine.
It's the opposite. You're building a foundation under a river. You sink a big inverted box, you pump compressed air into it to hold the water out, and men go down into it and excavate the riverbed from inside. The air pressure keeps the water where it is. Which means the men are working at two, three atmospheres, and when they come back up to normal pressure, the nitrogen that dissolved in their blood under pressure comes out of solution.
Bubbles.
Bubbles in the joints, in the spinal cord, in the lungs. It's a review by W.P. Butler in Undersea and Hyperbaric Medicine, two thousand four, on the Eads and Brooklyn Bridges, and it's the standard account of what this did to people.
And the fix didn't exist yet.
Staged decompression didn't exist as a discipline. Modern compressed-air work requires staged decompression and hyperbaric medicine. Then, you came up and you found out.
How routine was it?
Routine enough that it was just a thing that happened to sandhogs. That's the point. It wasn't a freak accident. It was an occupational hazard of the job, understood as such, and the men did it anyway because it paid.
And it still kills people now.
It still kills people now. New York City Water Tunnel Number Three killed twenty-three workers and one child over forty-six years of excavation. The child fell into an unsecured shaft.
Forty-six years.
And sandhogs working alongside tunnel boring machines can still suffer respiratory damage from dust exposure. The bends got regulated. The dust is a live issue.
So the modern era isn't clean either.
The modern era is much better. It isn't clean.
Now the growth, because I want the numbers.
Ridership. Two hundred and fifty-three million in nineteen hundred and one. Four hundred and forty-eight million in nineteen hundred and five. Seven hundred and twenty-five million in nineteen ten. One point three three billion in nineteen twenty. Two point oh five billion in nineteen thirty.
That's a city being rebuilt around a hole in the ground.
And the Dual Contracts, March nineteenth nineteen thirteen, Contracts Three and Four, with the IRT and the BRT, built or improved the majority of the system that exists today.
So the whole thing, the modern New York subway, is substantially a nineteen-thirteen construction programme.
Substantially, yes.
And now the part that answers Daniel's question.
New York's subway construction costs are now the highest in the world. Which has slowed expansion.
How slow?
The Second Avenue Subway was proposed in nineteen nineteen.
Nineteen nineteen.
Studied repeatedly. Started in nineteen seventy-two. Halted in nineteen seventy-five. Restarted in two thousand seven. First phase opened in two thousand seventeen.
Ninety-eight years. From proposal to a first phase of a subway line.
And I want to be careful here, because the obvious read is that safety killed it. That's not what the record says. The Second Avenue Subway was stopped in seventy-five because the city ran out of money. It's a fiscal crisis story, not a safety story.
That's the honest version. So let's actually do the accounting, because Daniel asked a specific question and I don't want to leave it hanging. Has the efficiency offset the loss?
The answer splits in two and I think that's the real finding. The first half: yes, unambiguously, on the things we stopped doing. Nobody is arguing for the bends as a productivity measure. Nobody wants the Fleet sewer in the works. The Metropolitan Mixture is not a heritage we should mourn.
And the second half?
The second half is that the numbers don't support the idea that safety is the main drag. If safety standards were the whole explanation, you'd expect a fairly uniform slowdown across countries with similar standards. You don't get that. New York is an outlier. It's the most expensive subway construction on earth, and Paris, with comparable labour protections and comparable environmental law, builds cheaper.
So it's not the rules. It's something about how the rules get applied and who's in the room.
And Paris gives you the counter-example on design. Bienvenüe showed that speed came from choosing cut-and-cover and designing for rapid repaving. That's a design decision, not a labour decision. Modern engineers have the same kind of levers. Tunnel boring machines, precast segmental lining, top-down construction. The lever exists. The question is whether anyone's allowed to pull it.
And there's a nice piece of evidence for that. Paris Line Fourteen, nineteen ninety-eight. First fully new metro line in sixty-three years, first fully automatic line with platform screen doors on every station from opening day.
And the Grand Paris Express, four new orbital lines, fifteen, sixteen, seventeen, eighteen. All planned as fully automated and fully accessible.
So Paris is still doing the thing Paris did. Plan the system, build the system.
And then there's the accessibility bill, which is the other side of the ledger. The historic Paris network is less than ten percent wheelchair accessible. Ranked worst in the world. Retrofitting is estimated at four to six billion euros, and RATP considers several stations impossible to retrofit.
Because they were built before the standard existed.
That's the same trade, one more time. The Victorians and the Edwardians built fast and cheap and left us a network that we now have to spend billions making usable, or write off.
So the bill arrives either way. You pay it in lives while you build, or you pay it in billions while you retrofit.
Or, in the case of the Second Avenue Subway, you pay it in a century of not having the line.
Right. And I want to bring it back to Jerusalem for a second, because Daniel framed it that way and I think it's the correct frame. The city we're sitting in is digging. Light rail, the J Net, high-rises, foundations going down everywhere. And the honest question isn't whether we should go back to hand-digging with navvies.
Obviously not.
The question is whether the coordination problem is actually a safety problem wearing a disguise. Because if it isn't, then we're paying the Victorian price without getting the Victorian speed.
Which is the worst of both. Can I put one number on that, because it's the cleanest version of the argument. New York spends more per kilometre of subway than any city on earth, and it does it under a safety regime that Paris also operates under. Same standards, wildly different outcomes. That gap is not safety. That gap is process.
So the modern drag is real and it isn't the thing we'd assume it is.
It's the thing Daniel suspected, I think. He wrote that it might be hard to envision a less coordinated process than Jerusalem's. And the answer from the history is that the Victorians weren't better coordinated. They were just less accountable. Nobody had to get consent. That's the difference.
And consent is expensive.
Consent is expensive and consent is correct. The trick is doing it without taking ninety-eight years.
I want to come back to one thing before we move on, because I think we skipped past it. The labour practices Daniel asked about specifically. Which of them would be illegal today?
Almost all of them, and it's worth listing rather than gesturing. Hand excavation in a trench with no shoring standard. Working in compressed air with no decompression protocol. No relief breaks on a shift. No dust control. No right to refuse. And the compensation question, because when a navvy was killed on the Metropolitan, there was no workers' compensation scheme to claim against.
So the family got nothing.
The family got whatever the contractor felt like giving them.
And the pace.
That's the honest link and I don't want to soften it. The speed came from the fact that a man was cheap and a delay was expensive. If you can replace a worker in a day and a sewer takes months to fix, you'll take the risk every time.
And that's the trade Daniel's asking about.
That's the trade. And the answer is that we made the right call and we're still paying for the way we made it.
Okay. I want to hand this over, because there's someone at this desk who has been sitting on something for about twenty minutes.
Hilbert: Twenty-two thousand cubic metres.
Sorry?
Hilbert: That's one section. Twenty-two thousand cubic metres of spoil out of one section of tunnel. I had a ledger. Muckaway coordinator, on a tunnel job. Not a miner. I tracked where the dirt went.
You were the spoil guy.
Hilbert: I was the spoil guy. Everyone talks about the workers and the engineers. Nobody talks about the dirt. You dig a tunnel, you get a mountain. It has to go somewhere, and it has to go somewhere that will take it, and it has to get there through a city, at night, on a permitted route, and every load has to be accounted for because if it isn't, someone dumps it in a field and you get a fine and a newspaper.
And that's the constraint.
Hilbert: That's the constraint. In London they dumped it in the river. In New York they used it to build up the shoreline. In Paris they piled it somewhere and called it a park. The dirt tells you how fast they were going. If you can put the spoil anywhere, you can dig as fast as you like.
So the pace was partly a waste-disposal question.
Hilbert: It was mostly a waste-disposal question. You can't dig faster than you can get rid of what you've dug. That's the whole thing. Everyone's talking about the bends and the smoke and the safety, and I agree with all of it, but the reason they could go that fast is that the dirt was free to get rid of. Ours wasn't.
What did it cost you?
Hilbert: Per load? I don't remember. I remember the ledger. I remember the columns. Section, cubic metres, destination, permit number, time out, time in. I could tell you how many loads went out on a Tuesday in a particular month and I probably still could, if I found it.
And the modern comparison is that you're moving the same dirt with a much longer paper trail.
Hilbert: Much longer. And that's not nothing. Half those loads went to a site that had to be remediated first, and the remediation had its own permit, and the permit had its own meeting. You dig a tunnel in a city now and you're really running a haulage company with a tunnel attached.
So when Daniel asks whether modern construction is slower, part of the answer is that we're not just building a tunnel. We're building a tunnel and a waste management operation and a public consultation.
Hilbert: And the waste management operation is the bit that has to finish before the tunnel can. You can't line a tunnel you haven't cleared. So the critical path runs through the dirt, not through the digging.
That reframes it, actually. Because we were treating the drag as a process problem, and you're saying the process problem has a physical unit.
Hilbert: It has a unit. It's a cubic metre. And you can measure it. If you want to know why a project is slow, ask where the spoil is going. If the answer is a long one, you've found it.
And in eighteen sixty the answer was the river.
Hilbert: In eighteen sixty the answer was the river, and the river didn't have a permit, and the river didn't have a meeting.
Did you enjoy the job?
Hilbert: I liked the ledger. I liked that at the end of a shift the numbers balanced. There's not a lot of work where you can say that.
Hm.
Hilbert: One correction, though. You said the drag isn't safety. It's not safety, but it's partly the same instinct. Somebody decided the dirt had to be accounted for. That's a good decision. It just costs money and it costs time, and nobody puts it on the poster.
That's fair.
Hilbert: And I'd still take the ledger over the river.
So let's land this properly, because I think Hilbert just handed us the actual answer and we should use it.
Go on.
The Victorian speed wasn't one thing. It was a bundle. Cheap labour, no consent, no compensation, and free disposal of waste. We've abolished three of those and I don't want any of them back. But the fourth one, the disposal, we replaced with a system that is correct and slow, and we never really looked at whether it could be correct and fast.
The honest answer to Daniel's question is that the efficiency has offset the losses on the things that mattered and hasn't offset them on the things that didn't. We're safer. We're not faster. And the gap between safer and faster is where the argument actually lives.
Here's the thing I'd put on the cutting-room floor, because it didn't fit anywhere and it's too good to lose. The London line was built with almost no ventilation because they believed it would run on smokeless locomotives. And when the smoke turned out to be a problem, the fix was to cut holes in the tunnel roofs, and the local authorities objected because the holes would frighten the horses.
Which is the whole episode in one anecdote.
It's the whole episode. Every generation solves the problem it can see and creates the one it can't.
The forward-looking version of this, for a city currently digging, is that the question isn't how to get back to Victorian speed. It's whether we can get the accountability without the century. Because New York has proved you can have modern standards and still take ninety-eight years, and Paris has proved you can have modern standards and still finish.
Which means the drag is a choice, not a law of nature.
It's a choice somebody made in a meeting.
That's the episode. Thanks to Hilbert Flumingtop, who produces this and apparently also knows exactly how much dirt came out of a tunnel. This has been My Weird Prompts. If you enjoyed this look at the hidden costs of progress, leave us a review and tell a friend. It helps more than you'd think. We'll be back soon.
My weird prompts dot com. See you then.
