Daniel sent us this one, and it's a question that's been quietly simmering beneath every car-free city announcement for the past decade. The headlines say a city is banning cars, and everyone pictures this utopia of wide-open plazas and bicycles gliding past cafés. But here's the actual question: when you remove the cars, you get back maybe six to ten meters of road width. That's it. And now you've got to divide that up between buses, trams, bikes, pedestrians, and fire trucks that still need to get through. So how are cities actually making those choices, and who's losing out?
This is the spatial politics nobody talks about. The anti-car movement is right about a lot of things, but there's this unspoken assumption that removing cars creates abundance. It doesn't. It creates a knife fight over about eight meters of asphalt. And the combatants aren't cars versus bikes. They're bikes versus buses, pedestrians versus trams, and everyone versus emergency services. I've been reading transport planning documents from about a dozen cities, and the pattern is the same everywhere: every mode has a minimum viable width, and those minimums add up to more than the available space.
The car-free city isn't really about removing cars. It's about choosing which non-car mode gets to win.
And different cities have chosen different winners. Let me give you the spatial constraint upfront so we're grounded in real numbers. A typical urban arterial street is twenty to thirty meters wide, building face to building face. Of that, you might have six to ten meters of actual lane width that cars currently occupy. You free that up. Now, a protected bike lane needs about two and a half to three meters in each direction. A dedicated bus lane needs three and a half meters. A tram corridor needs about seven meters for two tracks. Sidewalks need at least two and a half meters to handle pedestrian flow comfortably, and more in commercial areas. You add those up, and you're at sixteen to twenty meters of demand for ten meters of supply. Something gets cut.
That's before you even get to the stuff nobody wants to think about. Loading zones for businesses. If you want a mature tree canopy, the root zone alone needs about two meters of uninterrupted soil volume, which competes with every underground utility and every surface mode. So it's not even sixteen to twenty meters of demand. It's probably more like twenty-two to twenty-five once you factor in the non-transport uses of a street.
That's a crucial point, and it's one that street design guides are only starting to grapple with. The Dutch have a concept called "de complete straat," the complete street, and even their most advanced templates acknowledge that you can't fit everything. You have to choose a primary function for each street and let the other functions be secondary. A street that's primarily a mobility corridor looks different from a street that's primarily a destination. But the problem is that most car-free advocates want every street to be both.
Of course they do. So let's walk through who's making which cuts, because that's where the actual ideology of a city reveals itself.
And I want to use three cities as our framework here, because they represent three fundamentally different philosophies. Amsterdam put bikes first. Paris bet on trams and pedestrianization under Anne Hidalgo. And Bogotá went all-in on bus rapid transit, the TransMilenio system. Each of these choices created winners and losers, and the losers are instructive.
Start with Amsterdam. The default mental image of a car-free city.
Amsterdam has five hundred and fifteen kilometers of cycle paths. It's the global gold standard for cycling infrastructure. But what most people don't realize is that Amsterdam's bus network has been steadily degraded by that very success. The city's own transport report from twenty twenty-four showed that bus speeds dropped twelve percent on routes that share space with bike lanes. And there's a specific example that's become infamous in transport planning circles: the Weesperstraat redesign in twenty twenty-three. They removed car lanes, and they had to choose between a protected bike lane and a dedicated bus lane. They chose bikes. Bus route thirty-seven now takes eight minutes longer per trip.
On a bus route. That's not marginal. That's the difference between someone choosing the bus or giving up and buying a car.
Which is the irony that haunts all of this. You degrade the bus to prioritize bikes, and eventually the bus becomes so slow that people start driving again. Amsterdam hasn't hit that tipping point yet, but the tension is real. The tram system has its own conflict with bikes. Amsterdam's tram tracks are notorious for catching bike tires. If you've ever cycled there, you learn very quickly to cross tracks at a sharp angle or you're going over the handlebars. The city has experimented with rubberized track infill and dedicated bike crossings, but about forty percent of major tram routes still share corridors with bike lanes. There's simply not enough width to separate them.
The tram-bike interface is basically the handshake that breaks fingers.
That's exactly what it is. And it's not just a comfort problem. The Amsterdam fire department put out a report in twenty twenty-four showing that emergency response times are two to three minutes slower than in car-centric cities of comparable density. The reason is that fire trucks have to navigate streets where bike lanes, tram tracks, and narrowed vehicle lanes create a kind of obstacle course. They use bike lanes in emergencies, but that creates its own hazards.
I want to pause on that fire department report, because it gets at something deeper. When we talk about who loses in these spatial allocations, we tend to think in terms of convenience. The bus rider loses eight minutes. The cyclist has to take a longer route. But emergency response time isn't about convenience. It's about whether someone's house burns down completely or gets partially saved. It's about whether a stroke patient gets to the hospital within the golden hour. The stakes are fundamentally different, and I'm not sure the planning frameworks even have a way to weight that difference properly.
They don't. Cost-benefit analysis in transport planning typically uses value of time saved, which is an economic metric. But there's no widely accepted metric for "value of not dying in a fire." What you get instead is fire departments issuing reports like Amsterdam's, which are essentially bureaucratic screams into the void, and then the planning process continues as before because the political momentum behind bike infrastructure is so strong.
Amsterdam made its choice. Bikes are the primary mode, and buses, trams, and fire trucks all pay a tax on that choice. What did Paris do differently?
Paris under Anne Hidalgo took what I'd call the multi-modal gamble. The fifteen-minute city vision wasn't just about bikes. It was about trams, pedestrianization, and bikes all advancing simultaneously. Between twenty twenty and twenty twenty-five, Paris invested three hundred million euros in tram line extensions, the T3, T6, and T8 lines. They pedestrianized the Seine quays. They built hundreds of kilometers of bike lanes. And the result was that they created conflicts on all fronts simultaneously.
Spreading the pain around, rather than concentrating it on one mode.
And the most visible conflict has been between bikes and buses. On boulevards like Haussmann and Magenta, bike lanes ate into what were previously bus-only lanes. Bus speeds dropped twelve percent between twenty twenty and twenty twenty-five. On the Rue de Rivoli, they built a four-meter-wide bike lane, which is enormous by urban standards, and it forced the adjacent bus lane to narrow to two and a half meters. Standard Paris buses are two point five five meters wide. So the lane became functionally impassable.
They built a bus lane that buses can't fit in?
They had to deploy smaller minibuses on that route. The RATP, which is the Paris transport authority, had to buy a whole new fleet of narrower vehicles just for that corridor. That's not a design feature. That's a design failure that became an expensive procurement problem.
Paris's approach is essentially: let's do everything at once, and if the bus lane ends up being narrower than the bus, we'll buy narrower buses. That's a very French solution to a geometry problem.
And to be fair, the tram investment has paid off in terms of ridership. The T3 line now carries over two hundred fifty thousand passengers a day. But the bus degradation is real, and bus riders in Paris tend to be lower-income and more likely to be commuting from the banlieues. So there's an equity dimension here that doesn't get talked about enough. When you trade bus speed for bike lanes, you're often trading the mobility of outer-suburb commuters for the mobility of inner-city residents.
Which is the kind of observation that gets you uninvited from urbanist dinner parties. But let's dig into that equity point, because it's easy to say and harder to prove. How do we actually know that bus riders in Paris are lower-income? Is there data on that?
The Île-de-France mobility survey from twenty twenty-three broke down mode share by income decile. The bottom three income deciles account for nearly sixty percent of bus trips in the region but only about twenty percent of bike trips. Bike commuters in Paris skew heavily toward the top half of the income distribution. They're more likely to live in the city center, more likely to have flexible work schedules, and more likely to be male. Bus riders are the inverse on all three metrics. So when you slow down the bus to speed up the bike, you're effectively transferring travel time from wealthier inner-city residents to poorer suburban ones.
That's a regressive time tax. It's not a tax in euros, but it's a tax in minutes, and minutes are the one resource that's perfectly equally distributed. Everyone gets twenty-four hours. Taking eight minutes from a bus rider is taking a larger share of their disposable time than it would be from someone with more flexibility.
That's the argument that transport equity researchers have been making for years, but it doesn't cut through politically because the bike lobby is organized, visible, and concentrated in the urban core where political decisions get made. Bus riders are diffuse, less organized, and often live outside the municipal boundaries where they can't even vote in city elections.
What about Bogotá? That's a very different starting point.
Bogotá is fascinating because it approached car-free from the opposite direction. Most cities start with bikes and pedestrianization and then figure out transit later. Bogotá started with the TransMilenio bus rapid transit system, which now carries two point four million passengers a day. That's more than most metro systems in the world. It runs on dedicated busways down the median of major avenues. And cycling infrastructure was essentially an afterthought until twenty twenty, when the pandemic triggered the creation of five hundred and fifty kilometers of emergency bike lanes.
The hierarchy was inverted. BRT first, bikes second. And now they're competing for the same median strip.
And the conflict came to a head with the Carrera Fifteen BRT corridor expansion in twenty twenty-five. Carrera Fifteen, also called Avenida Caracas, is one of Bogotá's main north-south arteries. The expansion required removing three kilometers of emergency bike lanes that had been installed during the pandemic. The compromise was a raised cycle track built on the sidewalk rather than in the median. But that then conflicted with pedestrian flow, because Bogotá's sidewalks in that area are only about three meters wide.
They solved the bike-BRT conflict by creating a bike-pedestrian conflict. It's like a game of spatial whack-a-mole.
That's exactly what it is. And this is the pattern across all three cities. You can't actually solve the allocation problem. You can only move the conflict to a different interface. Amsterdam moved it to the bus-bike interface. Paris moved it to the bus-bike interface on different streets. Bogotá moved it to the bike-pedestrian interface. But the conflict itself is irreducible because the geometry is irreducible.
Which brings us to the wildcard you mentioned earlier. Because you can negotiate with bus riders and cyclists. You can't negotiate with a burning building.
This is where a lot of car-free idealism collides with physics. Fire trucks need a minimum clear width of three and a half meters. Ambulances need three meters. They need turning radii that don't exist on streets designed for bikes and pedestrians. And they need to get through without stopping for bollards or planters or café seating. Copenhagen has probably the most sophisticated solution to this. They installed emergency priority signals at a hundred and twenty intersections. When a fire truck or ambulance approaches, GPS triggers the traffic signals to give them a green wave. It reduced fire truck response times by about thirty seconds.
Thirty seconds is meaningful in a cardiac arrest. But what did they have to give up?
They had to remove pedestrian scramble phases at those intersections. A scramble phase is when all vehicle traffic stops and pedestrians can cross in any direction, including diagonally. It's great for pedestrian flow and safety. But it creates a delay for emergency vehicles because they have to wait for the full pedestrian cycle. Copenhagen decided that thirty seconds of emergency response time was worth more than diagonal pedestrian crossings. Reasonable trade-off, but it's still a trade-off.
Paris tried a different approach with the retractable bollards.
Paris installed retractable bollards on pedestrianized streets so that emergency vehicles could lower them and pass through. The problem is that during the twenty twenty-four heatwave, a bunch of them jammed. Mechanical systems fail in extreme temperatures. When a bollard is stuck in the up position, that street is closed to emergency vehicles until a maintenance crew arrives. That's not a theoretical risk. That's something that actually happened.
The Paris solution is great until it's forty degrees and someone's having a heart attack behind a jammed bollard. That's the kind of sentence that should make every city planner sweat.
Oslo learned this lesson the hard way. When they implemented their car ban in twenty nineteen, ambulance response times increased by four minutes. Four minutes is an eternity in emergency medicine. The fix was creating dedicated emergency corridors within pedestrian zones, two and a half meters wide, with no obstacles. But that required narrowing sidewalks and removing two hundred on-street parking spaces that had already been converted to pedestrian seating. The cost was a hundred and twenty million Norwegian kroner, about ten million euros.
The Oslo model is: budget fifteen to twenty percent of your pedestrianized space for emergency corridors, and accept that you're going to spend millions retrofitting something you should have designed in from day one.
That's the actionable lesson for any city planning a car-free zone now. Design emergency access in from the start. Don't retrofit it. The retrofit is always more expensive and more disruptive. Oslo's experience is now being studied by cities from Helsinki to Mexico City, because nobody wants to be the next city that has to explain a four-minute ambulance delay.
What I find striking about the Oslo case is that it wasn't a secret. The fire and ambulance services presumably raised concerns during the planning phase. So how did the city end up with a four-minute delay? Was it that the planners didn't listen, or that the emergency services didn't have the political weight to override the pedestrianization agenda?
From what I've been able to piece together from the post-implementation review, it was a combination of both. The emergency services did raise concerns, but they were treated as operational details to be solved later rather than as fundamental design constraints. The political timeline was driven by the car ban announcement, which had a fixed date. The technical work on emergency access was supposed to happen in parallel, but it got deprioritized because nobody wanted to be the person who said "we need to delay the car-free future because we haven't figured out where the fire trucks go." It's a classic case of political momentum overriding technical due diligence.
You mentioned ferry cities earlier. London and Sydney. How does water complicate this?
It adds an entirely new dimension of spatial competition. London's Uber Boat carries twelve million passengers a year on the Thames. That's real transport capacity. But ferry terminals need waterfront space for piers, waiting areas, and passenger circulation. And that waterfront space is also prime real estate for cycling infrastructure and pedestrian promenades. The Thames Path is only three meters wide in sections. You've got commuters trying to walk to work, tourists stopping to take photos, and cyclists trying to use it as a transport corridor, all in three meters.
That's narrower than a single lane of road. For three completely incompatible use cases.
The conflict came to a head in twenty twenty-four when the CS3 cycle superhighway, which runs along the Thames, was narrowed from four meters to two and a half meters to accommodate the Uber Boat terminal expansion. Cycling volumes on that segment dropped eight percent. So you've got a mode shift happening, but it's from bikes to, presumably, the Tube or buses. It's not mode shift from cars. It's mode shift between non-car modes.
Which is the thing that never makes it into the press release. The city announces a ferry terminal expansion, and nobody mentions that the bike lane just lost a meter and a half of width. Sydney handled this better, you said?
Sydney's twenty twenty-five Circular Quay redesign is instructive. They dedicated forty percent of the terminal space to bike parking. That's an unusually high allocation. Most ferry terminals give bike parking maybe five or ten percent of the space, if they think about it at all. The trade-off was that the pedestrian plaza area shrank by fifteen percent. So again, you're not creating space. You're reallocating it. Sydney chose to prioritize bike-ferry integration over pedestrian amenity. And that's a defensible choice if your goal is to make multi-modal journeys seamless. But it's still a choice with losers.
The pedestrian plaza didn't get smaller by accident. Someone made a spreadsheet and decided that bike parking was worth more than plaza square footage.
That spreadsheet exists somewhere in Transport for New South Wales, and I would love to see it. Because those allocation ratios are the closest thing we have to a revealed preference about what a city actually values, as opposed to what it says it values.
If we zoom out from the individual cities, is there a pattern in how cities make these choices?
Most car-free cities implicitly follow what planners call a reverse pyramid. Pedestrians at the top, then cyclists, then public transit, then emergency services, then freight and deliveries. But that pyramid is more of an aspiration than a design principle, and it breaks down as soon as you try to implement it. Because what does pedestrian priority actually mean when a fire truck needs to get through? The fire truck wins. It has to. So the real hierarchy is: emergency services first, whether you admit it or not, then everything else negotiates.
The pyramid is a nice piece of rhetoric that collapses on contact with a burning building.
And the more honest cities are starting to acknowledge this. Copenhagen's Red Route system, implemented in twenty twenty-three, gives emergency vehicles GPS-triggered signal priority on forty-five kilometers of bike-priority streets. It's an explicit acknowledgment that emergency access trumps cycling priority. But cyclists complain about near-misses at about twelve reported incidents per month, because a fire truck moving at speed through a bike corridor is inherently dangerous.
There's something almost darkly comic about a city building the world's best bike infrastructure and then having to run fire trucks through it. It's like building a beautiful garden and then driving a tractor through it twice a day.
That tension isn't going away. But there are some genuinely innovative approaches emerging that try to sidestep the zero-sum dynamic entirely. Barcelona's superblocks are the most interesting example. The Poblenou superblock, redesigned in twenty twenty-five, uses time-based allocation. Between seven and ten in the morning and four and seven in the evening, the lanes are bus-priority. Between ten and four and on weekends, they're bike-priority. Pedestrians have permanent priority on the interior streets. The allocation shifts by time of day using movable bollards.
The street has a schedule. It's a bus lane during rush hour and a bike lane during lunch. That's clever, but it also sounds like a signage and enforcement nightmare.
It relies on digital signage and, to be honest, a certain amount of social norming. Barcelona's transport culture is cooperative enough that it works. I'm not sure it would work in a city with more aggressive driving or cycling cultures. But the principle is sound: if you can't give everyone dedicated space all the time, give them dedicated space some of the time and make the schedule predictable.
I'm curious how the transition actually works in practice. If you're a cyclist using the lane at three forty-five in the afternoon, and at four o'clock the bollards retract and suddenly it's a bus lane, what stops you from getting trapped or, worse, hit?
The Barcelona system uses a fifteen-minute buffer period with flashing amber lights and audible warnings. The bollards don't retract all at once. They retract sequentially from one end of the corridor to the other, effectively sweeping the lane clear. It's not perfect. There have been incidents. But the injury rate is low enough that the system has survived political scrutiny. The bigger challenge has been compliance during the buffer period. Some cyclists treat the flashing lights as a suggestion rather than a requirement, which creates friction with bus drivers who are waiting to enter the lane.
Vienna took a different approach, right? Mode-specific streets.
Vienna's twenty twenty-four modal filter program designated certain streets as bus-only, others as bike-only, with pedestrian streets connecting them. The idea is that instead of trying to fit every mode on every street, you create a network where each street does one thing well. If you're a cyclist, you use the bike streets and accept that you might have to go a block out of your way. If you're a bus, you stay on the bus streets and don't have to share with bikes. The trade-off is that journeys become slightly less direct for everyone, but the dedicated space means they're faster and safer.
It's the transport equivalent of a zoning code. This street is zoned for bikes. That street is zoned for buses. Mixed-use streets create conflict, so separate them.
It works if the network is dense enough that the detours are trivial. In a city like Vienna, with a well-connected grid, adding a block to your journey is negligible. In a city with a sparser street network, it might not work. Tokyo is pushing this even further with vertical stacking. Their twenty twenty-six pilot on the Yamanote line corridor is testing elevated bike lanes above tram tracks. It's expensive, but it essentially eliminates the spatial conflict by adding a third dimension.
Stacking transport modes like a layer cake. Bikes above trams above pedestrians at grade. It's the only solution that actually creates new space rather than reallocating existing space.
If the Tokyo pilot works, it could be the model for dense cities everywhere. The cost is the obvious barrier. Elevated structures are orders of magnitude more expensive than paint and bollards. But if you're a city like New York or London where land values are astronomical, the cost of elevated bike infrastructure might actually be lower than the opportunity cost of dedicating surface space to bikes instead of pedestrians or buses.
This is where the conversation gets uncomfortable for the tactical urbanism crowd. Paint and planters can only get you so far. At some point, you need concrete and steel and serious money.
That's the thing I wish more car-free advocates would acknowledge. The easy wins are behind us. The first wave of bike lanes and pedestrian plazas used space that was underutilized or could be taken from cars with minimal disruption. The next wave is going to require hard choices between modes that are all legitimate and all have political constituencies. And the cities that navigate this well are going to be the ones that are transparent about the trade-offs, not the ones that pretend everyone can have everything.
If you're a listener in a city that's planning car-free zones, what should you actually do with all this?
First, understand that your city has already made a choice about which mode it prioritizes, whether it admits it or not. Amsterdam chose bikes. Bogotá chose BRT. Paris is trying to do everything and struggling with the bus-bike conflict. If you're moving to or within a car-free city, know which mode is the primary one and plan your life around it. Don't be the person who moves to Amsterdam and complains that the bus is slow. The bus is slow because the city chose bikes. That's the deal.
Second thing: emergency access. If your city is planning a car-free zone, go to the public consultation and ask the specific question. What is the minimum clear width for emergency vehicles on every street? Has the fire department signed off? Has the ambulance service done a response time simulation? Oslo's four-minute delay didn't have to happen. It happened because nobody asked the question until after the bollards were in the ground.
Third, ask for the mode hierarchy document. Amsterdam publishes something called the Street Space Allocation Plan. It's a public document that shows, street by street, how much width is allocated to each mode. If your city can't or won't produce something like that, they haven't actually done the planning. They're just painting bike lanes and hoping for the best.
The allocation ratio is the honesty test. If a city says it's prioritizing pedestrians but can't tell you what percentage of street width goes to sidewalks versus bike lanes versus bus lanes, they're not prioritizing anything. They're improvising.
Improvisation is how you end up with a bus lane that's narrower than the buses. Which brings me to the open question I keep coming back to. What happens when autonomous vehicles enter this picture? Several cities, including Helsinki and Singapore, are already planning for AV-dedicated lanes. But is an autonomous vehicle a car that should be banned from car-free zones, or is it public transit that should be prioritized? If a self-driving shuttle carries eight people and runs on a fixed route, how is that different from a bus?
It's different because someone owns it. And the ownership question is going to be the political fault line. A municipally owned autonomous shuttle is public transit. A privately owned autonomous taxi is a car with better PR.
If AV-dedicated lanes start appearing in car-free zones, we're going to see a version of the same spatial conflict we've been discussing, but with a new combatant. The AV lane will need three meters of width, same as any vehicle lane. That's three meters that currently belongs to bikes or buses or pedestrians. The geometry hasn't changed just because the vehicle doesn't have a driver.
The autonomous vehicle is just the next mode asking for its three meters. And the street is still only twenty meters wide. The fundamental problem hasn't changed since the Romans laid out the first grid. Space is finite, and every new mode creates a new claimant. Car-free doesn't mean conflict-free. It just means the conflicts are between modes that are all, in their own way, the good guys.
On that cheerful note, I think we've earned a palate cleanser. Hilbert, what have you got for us today?
And now: Hilbert's daily fun fact.
Hilbert: In the nineteen thirties, the American anthropologist Leland Locke studied a single surviving quipu from the Inca Empire that recorded tax obligations across seventeen villages using a base-ten positional system of knots, proving that the Andean civilizations had developed sophisticated accounting independent of any written language — and that particular quipu, catalogued as AS187 at the American Museum of Natural History, remains the most complete pre-Columbian fiscal document ever recovered from the Americas.
A single tax return made of string. There's something almost poetic about that.
At least it couldn't crash your computer.
This has been My Weird Prompts. If you enjoyed the episode, we'd love a review wherever you listen. It helps other people find the show. I'm Herman Poppleberry.
I'm Corn. Back next week.
