Interesting and thought-provoking article. It brought to mind the early planning of city builder Brigham Young in the American West. When he and the first pioneers laid out Salt Lake City, Utah in the late 1840s, they followed the Plat of Zion model established by Joseph Smith, designing streets that were 132 feet wide—broad enough for a team of oxen and a wagon to turn completely around without unhitching. This remarkable foresight, both practical and visionary, allowed for future growth, transport, and civic design well beyond the settlers’ immediate needs.
That same kind of long-range thinking—reserving space for movement, connection, and future innovation—is precisely what developing cities need today. Infrastructure planned with that balance of utility and imagination becomes a legacy, not a limitation.
I cite their work a couple times near the end, but it's definitely buried in there. As an aside, their initial intervention in Ethiopia focused on both 1) the total quantity of roads and 2) reducing circuity via grids. I think the latter is important, and something I've been meaning to write about separately, but from an initial analysis, my sense was the effect size of “more roads” was much larger than that of reducing circuity, at least outside of particularly upsetting cases.
Actually, this is going to be addressed in the not so distant future.
Autonomous cars are going to do away with personal auto ownership. People kill 41000+ other people in auto accidents in the USA alone, each and every year! As autonomous cars like Waymo and in 2026, Uber's anticipated entry into this arena, expand and proliferate, it will be inevitable that humans will be prohibited from driving, first on highways and then on city streets.
People will simply use an app to call for a car whenever they need it. With fleets logically positioned geographically, a ride may never be more than 10 minutes away.
Once humans are prohibited from driving, there will not be any need for street parking and therefore, all of the then unused street space formerly used for parking can be redeveloped.
Have you visited the ultra crowded streets of cities such as Mumbai where car, trucks, tuktuks bicycles, pedestrians and animals share roads full of obstructions? Autonomous vehicles will never be effective in these situations.
India's first priority should be instituting some form of mandatory birth control to significantly reduce their excessive population. This should have been done ages ago. AI/robots are going to be doing the majority of work in the world within 10-20 years. The last thing the world needs now is more people.
Therefore, due to the overpopulation numbers in countries like India and Bangladesh, people killed in such places due to auto accidents are of no major concern and the deployment of autonomous vehicles will likely come much later than elsewhere.
There are so many things within 20cms (8 inches) of a moving vehicle that a vehicle is likely to be repeatedly “paralysed” during any journey and not make as much progress as a vehicle piloted by a human
Because such close manoeuvres require human 2 way interaction and intuition . The human/car interface would be much slower and narrower in scope. In many situations the autonomous car would not have enough information to proceed safely so would remain stationary. This comment is based on walking in cities in India, China and Japan and two years living and driving in Sri Lanka
Autonomous cars will all be networked and connected. They will work together and coordinate among each other. The fly in the ointment, as always, are humans who get in the way.
This is why I say that the in first world countries, such as the USA, humans will be prohibited from driving on highways initially and eventually, also on city streets, leading to an Uber/Lyft fleet transportation model where people will not own cars but will call for service when they have a need for one.
3rd world countries, such as India, are a different story. Less people in these countries will make the streets less crowded and more navigable by autonomous cars. And also allow such countries to pull themselves up to 2nd world and first world levels.
India is already below replacement rate country-wide, below developed countries such as the US in some southern states (with accompanying demographic problems/immigrant labour controversies). Why do you think overpopulation is a problem?
(Moreover, are you really saying that life is so cheap in India that policies to reduce mortality don’t happen, or aren’t worth talking about? People vote, people don’t like being run over…)
Roads for what is the important bit here. Wide enough roads that put the private car absolutely dead last in priority, yes. But we should be very wary of following a big, wide road model that doesn’t really work for cities like Houston or Los Angeles … and once built is very hard to work back from.
Wider roads for walking, cycling, and transit, sure, have at it!
LA and Houston have a much better setup economically than, say, Mumbai, though. And it is very difficult to use public space for walking, cycling, or railways if you don't have the public space in the first place.
They might have a better set up economically just now, but will they continue to? Affording the upkeep of that sprawling space is a big problem! The trajectory of Mumbai is rather different too.
And I’d argue Manhattan isn’t the ideal. But Paris might be. Roads wide enough, but without consuming the city to traffic, though that required a radical adjustment to make work in Paris remember.
We’ve already seen the physical, social and economic scars that road widening programs have done to lots of cities too.
Barcelona's Eixample is demonstrably a more successful neighbourhood economically than the Gothic Quarter, and it extends to 'can you easily build a subway there?'
Wut. The headline here is totally wrong. Wide streets are bad, actually. Grids are great – and that seems to be what the article is actually about. Grids of narrow streets that are close together. Manhattan’s grid actually sucks because while the east west streets, don’t have very much space between them, the north south streets are way too far apart. The grid should make squares, not rectangles.
Or consider how many great cities: Manhattan, Berlin, Milan, Barcelona, and Madrid have largely wide streets.
The ideal is a mixture of streets whose purpose is mainly economic benefits (wide ones to move people around to maximise agglomeration) and streets whose purpose is mainly consumption (narrow ones).
Without mechanised transport, the maximum labour market is constrained substantially, and we have to live at uncomfortable overcrowding densities. One of the marvels of the modern world is how we created paved roads (and thus speedier waggons), canals, navigable rivers, railways, horse-drawn then electric trams, omnibuses, bicycles and the car.
All of these allow us to create larger labour markets, meaning a deeper division of labour, and to allow even the poorest to consume vastly more living space than they did 100 years ago.
It's very important to start with that point. Circulation is about moving people around more quickly, yes, but there is a direct correspondence between movement speed and overall economic output and living standards.
It's a separate question as to whether setting aside generous rights of way for transport means that we have to live, work, and play next to wide arterial roads. I would argue that we don't, and I would argue that the examples we hold up in the piece (especially Madrid and Barcelona, with particularly large shares set aside for roads) are pleasant balances, which developing countries should emulate.
Right - exactly, your last paragraph. The point of the article is that it’s important to have - most generally - a lot of space for roads, and in implementation that seems to mean a lot of roads. It’s also necessary to have a good grid. Not necessarily wide roads! “The developing world needs more roads.”
I think that, like Manhattan, Madrid, and Barcelona, developing countries need to lay out wide rights of way in new city areas, that they should use initially for roads. In a very successful world, they will use many of these for things other than private car transport.
The constant need for movement, huh. "The will to live!" Move, move, move, money, money, money. Leave no legacy, just move. How about this - look at the best cities in the world - emulate them. Stop this experimental transportation focused city planning. It's awful, can't you understand this by now? Build places worth keeping around.
One negative of wide streets that will bite extra hard in Africa is that you lose natural shade and everything is exposed to the tropical sun. Whoever lives in the houses whose front is turned towards the street, will be baking.
This can be mitigated by air conditioning, if your grid is reliable, but the expenses for power are non-trivial.
Meanwhile in London, the Greater London Council (Sadiq Khan) and Local Boroughs are closing roads and otherwise cutting capacity in very material amounts. Local transport is very good but at or beyond capacity during rush hour. My eldest daughter leaves for work shortly after 7am and often can't get a Line Bike.
Interesting and thought-provoking article. It brought to mind the early planning of city builder Brigham Young in the American West. When he and the first pioneers laid out Salt Lake City, Utah in the late 1840s, they followed the Plat of Zion model established by Joseph Smith, designing streets that were 132 feet wide—broad enough for a team of oxen and a wagon to turn completely around without unhitching. This remarkable foresight, both practical and visionary, allowed for future growth, transport, and civic design well beyond the settlers’ immediate needs.
That same kind of long-range thinking—reserving space for movement, connection, and future innovation—is precisely what developing cities need today. Infrastructure planned with that balance of utility and imagination becomes a legacy, not a limitation.
It's worth knowing the Urban Expansion Planning work of Patrick Lamson-Hall, now at Africa Urban Lab -https://www.aul.city/
Patrick's work is what initially introduced me to this topic actually!
LOL!! I skimmed through looking for a reference, but there were too many links to follow :-)
I cite their work a couple times near the end, but it's definitely buried in there. As an aside, their initial intervention in Ethiopia focused on both 1) the total quantity of roads and 2) reducing circuity via grids. I think the latter is important, and something I've been meaning to write about separately, but from an initial analysis, my sense was the effect size of “more roads” was much larger than that of reducing circuity, at least outside of particularly upsetting cases.
Actually, this is going to be addressed in the not so distant future.
Autonomous cars are going to do away with personal auto ownership. People kill 41000+ other people in auto accidents in the USA alone, each and every year! As autonomous cars like Waymo and in 2026, Uber's anticipated entry into this arena, expand and proliferate, it will be inevitable that humans will be prohibited from driving, first on highways and then on city streets.
People will simply use an app to call for a car whenever they need it. With fleets logically positioned geographically, a ride may never be more than 10 minutes away.
Once humans are prohibited from driving, there will not be any need for street parking and therefore, all of the then unused street space formerly used for parking can be redeveloped.
Have you visited the ultra crowded streets of cities such as Mumbai where car, trucks, tuktuks bicycles, pedestrians and animals share roads full of obstructions? Autonomous vehicles will never be effective in these situations.
India's first priority should be instituting some form of mandatory birth control to significantly reduce their excessive population. This should have been done ages ago. AI/robots are going to be doing the majority of work in the world within 10-20 years. The last thing the world needs now is more people.
Therefore, due to the overpopulation numbers in countries like India and Bangladesh, people killed in such places due to auto accidents are of no major concern and the deployment of autonomous vehicles will likely come much later than elsewhere.
My point was based on walking along such streets
There are so many things within 20cms (8 inches) of a moving vehicle that a vehicle is likely to be repeatedly “paralysed” during any journey and not make as much progress as a vehicle piloted by a human
If a human can navigate what sounds like an obstacle course, why would you believe that a machine could not do the same (or better)?
Because such close manoeuvres require human 2 way interaction and intuition . The human/car interface would be much slower and narrower in scope. In many situations the autonomous car would not have enough information to proceed safely so would remain stationary. This comment is based on walking in cities in India, China and Japan and two years living and driving in Sri Lanka
Autonomous cars will all be networked and connected. They will work together and coordinate among each other. The fly in the ointment, as always, are humans who get in the way.
This is why I say that the in first world countries, such as the USA, humans will be prohibited from driving on highways initially and eventually, also on city streets, leading to an Uber/Lyft fleet transportation model where people will not own cars but will call for service when they have a need for one.
3rd world countries, such as India, are a different story. Less people in these countries will make the streets less crowded and more navigable by autonomous cars. And also allow such countries to pull themselves up to 2nd world and first world levels.
India is already below replacement rate country-wide, below developed countries such as the US in some southern states (with accompanying demographic problems/immigrant labour controversies). Why do you think overpopulation is a problem?
(Moreover, are you really saying that life is so cheap in India that policies to reduce mortality don’t happen, or aren’t worth talking about? People vote, people don’t like being run over…)
Roads for what is the important bit here. Wide enough roads that put the private car absolutely dead last in priority, yes. But we should be very wary of following a big, wide road model that doesn’t really work for cities like Houston or Los Angeles … and once built is very hard to work back from.
Wider roads for walking, cycling, and transit, sure, have at it!
LA and Houston have a much better setup economically than, say, Mumbai, though. And it is very difficult to use public space for walking, cycling, or railways if you don't have the public space in the first place.
Manhattan is the ideal. Huge wide spaces.
They might have a better set up economically just now, but will they continue to? Affording the upkeep of that sprawling space is a big problem! The trajectory of Mumbai is rather different too.
And I’d argue Manhattan isn’t the ideal. But Paris might be. Roads wide enough, but without consuming the city to traffic, though that required a radical adjustment to make work in Paris remember.
We’ve already seen the physical, social and economic scars that road widening programs have done to lots of cities too.
Barcelona's Eixample is demonstrably a more successful neighbourhood economically than the Gothic Quarter, and it extends to 'can you easily build a subway there?'
Wut. The headline here is totally wrong. Wide streets are bad, actually. Grids are great – and that seems to be what the article is actually about. Grids of narrow streets that are close together. Manhattan’s grid actually sucks because while the east west streets, don’t have very much space between them, the north south streets are way too far apart. The grid should make squares, not rectangles.
There are virtues to narrow streets, but they are worse for circulation. See Smeed's 1963 paper https://x.com/bswud/status/1904838270879510890.
Or consider how many great cities: Manhattan, Berlin, Milan, Barcelona, and Madrid have largely wide streets.
The ideal is a mixture of streets whose purpose is mainly economic benefits (wide ones to move people around to maximise agglomeration) and streets whose purpose is mainly consumption (narrow ones).
“Circulation” is a synonym here for “cars go fast”. But we don’t want cars to go fast near where people live, walk, exist.
Without mechanised transport, the maximum labour market is constrained substantially, and we have to live at uncomfortable overcrowding densities. One of the marvels of the modern world is how we created paved roads (and thus speedier waggons), canals, navigable rivers, railways, horse-drawn then electric trams, omnibuses, bicycles and the car.
All of these allow us to create larger labour markets, meaning a deeper division of labour, and to allow even the poorest to consume vastly more living space than they did 100 years ago.
It's very important to start with that point. Circulation is about moving people around more quickly, yes, but there is a direct correspondence between movement speed and overall economic output and living standards.
It's a separate question as to whether setting aside generous rights of way for transport means that we have to live, work, and play next to wide arterial roads. I would argue that we don't, and I would argue that the examples we hold up in the piece (especially Madrid and Barcelona, with particularly large shares set aside for roads) are pleasant balances, which developing countries should emulate.
Right - exactly, your last paragraph. The point of the article is that it’s important to have - most generally - a lot of space for roads, and in implementation that seems to mean a lot of roads. It’s also necessary to have a good grid. Not necessarily wide roads! “The developing world needs more roads.”
I think that, like Manhattan, Madrid, and Barcelona, developing countries need to lay out wide rights of way in new city areas, that they should use initially for roads. In a very successful world, they will use many of these for things other than private car transport.
The constant need for movement, huh. "The will to live!" Move, move, move, money, money, money. Leave no legacy, just move. How about this - look at the best cities in the world - emulate them. Stop this experimental transportation focused city planning. It's awful, can't you understand this by now? Build places worth keeping around.
Which do you consider to be the best cities in the world? The piece suggests Manhattan and Madrid are places to learn from. Do you disagree?
One negative of wide streets that will bite extra hard in Africa is that you lose natural shade and everything is exposed to the tropical sun. Whoever lives in the houses whose front is turned towards the street, will be baking.
This can be mitigated by air conditioning, if your grid is reliable, but the expenses for power are non-trivial.
Meanwhile in London, the Greater London Council (Sadiq Khan) and Local Boroughs are closing roads and otherwise cutting capacity in very material amounts. Local transport is very good but at or beyond capacity during rush hour. My eldest daughter leaves for work shortly after 7am and often can't get a Line Bike.
Excellent. Many thanks!
China is exploiting this big time. They're creating 19 conurbations that raise local productivity 15%. Are standalone cities obsolete?
https://herecomeschina.substack.com/p/are-standalone-cities-obsolete-chinas?r=16k
The lift is the king. 2x1