I had a suspicion that there were more idiots on the road after and during covid. good to see this reflected in the data, but sad to know that it's actually true. no matter where I go in the USA I see people speeding easily 50% over the limit, running red lights, blowing through stop signs. it's ridiculous.
given how many people die I'm surprised government's having made safety technology mandatory. things like toyota safety sense are pretty effective - you can check on youtube. people will place random dummys in front of the car and it stops pretty accurately.
Pedestrians are hit by cars in the USA because the roads are not designed for non-car users. This is exacerbated by distracted driving, drunk driving, and recent car design changes like higher hood heights but the root of it is poorly designed roads which don’t consider pedestrians’ needs.
It is literally 20 times as many people killed every year as are killed in mass shootings but you can't get anyone to care at all about it. Blame the victim, did you see what they were wearing?
This kind of problem is exactly what statistics is designed to do, and it makes me a bit sad that we are left with a bit of a shoulder shrug. It's absolutely possible to do a much better job at disentangling possible causes here with something as simple as a multilevel regression. (Although ok, proper causal inference would be more work).
Also if you look at they way the cars are designed in the US compared to Europe, the hoods of the cars are much higher, and not designed to prevent injury in the event of a car -> pedestraion incident
- poor visibility in modern cars due to rollover protection
- touch screens and touch controls in cars
- general proliferation of controls in cars
- smart phones & smart phone addictions
- higher vehicle belt lines are better for vehicle --> vehicle impacts but worse for vehicle --> pedestrian impacts
- poor pedestrian infrastructure, sidewalks, crosswalks, etc.
We really really really really like our cars/trucks/SUVs in the US and have agreed that about 30,000 to 40,000 people a year will die so that we can keep driving the way we do.
It’s the price we pay for the way we choose to live.
my impression was not that most people like their SUVs, they're just don't trust other people and don't follow the conventions. While on the micro level Americans follow the conventions much more seriously than Europeans (starting from tips and down to formal office rules) and love spending time discussing those conventions, on the higher level, with risk or high stakes involved, all the conventions go out of window, and the people fall back to guns, litigators, SUVs, suburban houses in the middle of nowhere, and other ways of self-isolation and atomization.
Soccer vs american football is another visible example.
At first I thought maybe the number of pedestrian journeys have gone up, but that appears to be declining (leading to even more concern as to why deaths have gone up).
Where I live they will randomly build a bike path for a mile on the side of the road. But then it just ends. There's not a sense of how it could be built to connect people to the places they need to go. It's random and ad-hoc. Then people say "its pointless to build bike / ped infrastructure, nobody uses it!"
Great analysis - though I can't help but notice that 2009 is right when smartphones really took off (iPhone in 2007, Android in 2008, then mass adoption). The data showing accidents getting more deadly rather than more frequent actually makes sense if you combine two factors: phones causing more distracted driving incidents, plus our bigger American vehicles turning what would be injuries elsewhere into deaths. That could explain why it's US-specific - other countries probably have the same phone distraction problem, but their smaller cars mean less fatal outcomes. The distraction data might be weak simply because people don't admit they were on their phone after killing someone, but sometimes the obvious answer deserves more weight than we give it.
This morning while jogging in the US I came to an intersection. Green lights and walk on in my direction. A car approaching from my left had a red light, the driver glanced to his left and without stopping or looking in my direction, turned right across my path. I expected this of course, so avoided being run over. If I wasn't watching for this, it likely would be a different outcome.
So why do so many pedestrians get killed in the US? The two main reasons to me are: 1. Drivers don't look for pedestrians, and 2. pedestrians expect drivers to follow rules.
Another contributing factor is of course the huge vehicles that crush people with drivers barely noticing...
> So why do so many pedestrians get killed in the US? The two main reasons to me are: 1. Drivers don't look for pedestrians, and 2. pedestrians expect drivers to follow rules.
You think that isn't the same everywhere? I've got some news: in every country there are parents distracted by kids fighting in the back seat, and in every country pedestrians walking into light poles while on the phone is a running joke. Also: the USA has managed to export it's love for large cars to most countries. Here in Australia we call large SUV's shopping trolleys.
Despite this, if you look at the graphs in the article, you will see most countries have managed to drive down pedestrian deaths. Except the US, where the curve trends up. The reason is pretty straight forward, and has nothing to do with the cars, the attitudes of drives or pedestrians. Hell, you can even ask an AI what it is, and you will get a reasonable answer:
[Countries] have historically managed to drive down pedestrian deaths due to motor vehicle accidents primarily by adopting the Safe System approach, which includes elements of Vision Zero, a long-term goal of zero road fatalities and serious injuries.
This approach focuses on creating a road system that is safe for all users, particularly vulnerable ones like pedestrians, by managing speed, designing safer infrastructure, and ensuring safe vehicles and road user behavior.
The AI drones on and on, listing the many changes to road design and rules that caused the drop. This is not rocket science. Everybody can do it, and it's trivial to find out what needs to be done. What the USA lacks is a political system that can deliver it.
To me the reason is the right-on-red rule. I still find it insane, and I have been in Canada for 9 years.
In my driving classes, I have been clearly explained that a right-on-red must be treated like a stop sign and that to turn, there needs to be two lanes free of cars: the one you are getting into and the next one (if one lane is available,this doesn't apply).
Many,many drivers treat the red light like a green light for turning right and that's the root of the issue.
One obvious direction not explored in the article: looking not just at the type on the vehicle involved in the deadly collision, but also the actual model and its geometry.
It's not just that SUVs are deadlier than sedans.
It's also that the sedans are becoming taller, wider, heavier — and deadlier.
The article says that blunt fronts are what makes a collision more likely to result in a death. Well compare a 2000 Camry to a 2025 one then on that metric.
To test this hypothesis, we need to look at all accidents where a pedestrian was hit — and see a breakdown on whether it resulted in a fatality, by vehicle and road type.
Another thing the article doesn't consider is that the speed limits have increased across the US, and where they haven't, the enforcement is not necessarily there (cough Bay Area cough).
Solutions like lane diet (or engineering cities for anything other than automobiles) never became popular.
The outcome is inevitable.
_____
TL;DR: bigger, fatter cars going faster kill more people.
I don’t buy the distraction numbers. I see people on their phones constantly while driving, despite laws against it. It’s also impossible to really prove anything after the fact, as the article touches on. The graph shows a massive increase in “distraction not reported,” which to me just sounds like the driver didn’t choose to incriminate themself.
The spike started in 2010, which is when 4G was rolling out, Instagram launched, Facebook was already big, and social media in your pocket was becoming an addictive reality. Before this, there wasn’t a lot to do on a smartphone while driving.
in other countries at an intersection with a stoplight... is it normal for the light to turn green (allowing for a right turn) while a crosswalk also simultaneously activates to cross that path?
this always feels strange to me in some US cities... why would you give a car a green light to turn into an active crosswalk? I've been honked at by drivers like I'm doing something wrong while having a cross signal
Road and street design is the main difference with Europe. A lot of work has been put to make streets safer for everybody.
That fact alone acts as a multiplier for everything else. The cars are bigger in the USA? The bad streets make it way worse. People are distracted at the phone? The bad street design makes it more deathly.
Fix USA's streets and towns and all kinds of deaths will be decreased. It is the most important factor.
Anecdotally, on my few visits to the US (NY and Colorado), the driving I saw was absolutely atrocious compared to Europe. People were swerving and failing to stay in their lanes on Interstates, everyone was speeding well above the limit (everyone speeds on Motorways, but it seems it's taken to another level in the US). Then you have turn right on red meaning drivers just don't care and turn regardless. Then you have everyone driving massive fucking trucks where you can't see anything from inside.
It seemed every morning I got up and turned on the hotel TV, there was another news about some crash on the Interstate that morning
As an enthusiast of traffic engineering, the most surprising thing in US is how hard is for the engineers to handle so distinct zoning laws according to it's county/city/state, and urbanistic planning in several big/medium cities is more centered on _giving preference to the cars_ instead of _keep the cadency and flow of the cars_.
Not saying it's good or bad, but for instance, in some counties it's way simpler to have a parking lot without any traffic buffer area at the entrance than to get an approval for a roundabout to reduce electronic traffic coordination in feeder roads.
Even simple things like pedestrian passages that do not have any contact with the road (elevated passages or underground passages) are very hard to find in the US.
I really would like to know one day what kind of design philosophy the traffic engineering field follows with so much compromises.
I wish they wouldn't just focus on deaths. The difference between being killed and having your body wrecked is pretty small. I'm curious to know what the numbers look like if we considered some less extreme interpretation of taking someone's life.
59 comments
[ 3.6 ms ] story [ 59.6 ms ] threadgiven how many people die I'm surprised government's having made safety technology mandatory. things like toyota safety sense are pretty effective - you can check on youtube. people will place random dummys in front of the car and it stops pretty accurately.
This kind of problem is exactly what statistics is designed to do, and it makes me a bit sad that we are left with a bit of a shoulder shrug. It's absolutely possible to do a much better job at disentangling possible causes here with something as simple as a multilevel regression. (Although ok, proper causal inference would be more work).
They checked so many things I'm surprised it didn't match something just by accident (it's still a fun exercise :)) mostly just teasing)
We really really really really like our cars/trucks/SUVs in the US and have agreed that about 30,000 to 40,000 people a year will die so that we can keep driving the way we do.
It’s the price we pay for the way we choose to live.
Soccer vs american football is another visible example.
Or a raw increase in pedestrians on urban roads? Maybe people are more willing to go on a walk at night in the city these days?
So why do so many pedestrians get killed in the US? The two main reasons to me are: 1. Drivers don't look for pedestrians, and 2. pedestrians expect drivers to follow rules.
Another contributing factor is of course the huge vehicles that crush people with drivers barely noticing...
You think that isn't the same everywhere? I've got some news: in every country there are parents distracted by kids fighting in the back seat, and in every country pedestrians walking into light poles while on the phone is a running joke. Also: the USA has managed to export it's love for large cars to most countries. Here in Australia we call large SUV's shopping trolleys.
Despite this, if you look at the graphs in the article, you will see most countries have managed to drive down pedestrian deaths. Except the US, where the curve trends up. The reason is pretty straight forward, and has nothing to do with the cars, the attitudes of drives or pedestrians. Hell, you can even ask an AI what it is, and you will get a reasonable answer:
The AI drones on and on, listing the many changes to road design and rules that caused the drop. This is not rocket science. Everybody can do it, and it's trivial to find out what needs to be done. What the USA lacks is a political system that can deliver it.In my driving classes, I have been clearly explained that a right-on-red must be treated like a stop sign and that to turn, there needs to be two lanes free of cars: the one you are getting into and the next one (if one lane is available,this doesn't apply).
Many,many drivers treat the red light like a green light for turning right and that's the root of the issue.
It's not just that SUVs are deadlier than sedans.
It's also that the sedans are becoming taller, wider, heavier — and deadlier.
The article says that blunt fronts are what makes a collision more likely to result in a death. Well compare a 2000 Camry to a 2025 one then on that metric.
To test this hypothesis, we need to look at all accidents where a pedestrian was hit — and see a breakdown on whether it resulted in a fatality, by vehicle and road type.
Another thing the article doesn't consider is that the speed limits have increased across the US, and where they haven't, the enforcement is not necessarily there (cough Bay Area cough).
Solutions like lane diet (or engineering cities for anything other than automobiles) never became popular.
The outcome is inevitable.
_____
TL;DR: bigger, fatter cars going faster kill more people.
The spike started in 2010, which is when 4G was rolling out, Instagram launched, Facebook was already big, and social media in your pocket was becoming an addictive reality. Before this, there wasn’t a lot to do on a smartphone while driving.
That's over 3_000_000 people in the past 100 years.
https://en.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in...
(corrected, thanks!)
this always feels strange to me in some US cities... why would you give a car a green light to turn into an active crosswalk? I've been honked at by drivers like I'm doing something wrong while having a cross signal
That fact alone acts as a multiplier for everything else. The cars are bigger in the USA? The bad streets make it way worse. People are distracted at the phone? The bad street design makes it more deathly.
Fix USA's streets and towns and all kinds of deaths will be decreased. It is the most important factor.
It seemed every morning I got up and turned on the hotel TV, there was another news about some crash on the Interstate that morning
Not saying it's good or bad, but for instance, in some counties it's way simpler to have a parking lot without any traffic buffer area at the entrance than to get an approval for a roundabout to reduce electronic traffic coordination in feeder roads.
Even simple things like pedestrian passages that do not have any contact with the road (elevated passages or underground passages) are very hard to find in the US.
I really would like to know one day what kind of design philosophy the traffic engineering field follows with so much compromises.