A terrific / terrifying example for teaching externalities.
Folks choose heavier cars because they're less likely to die in a crash (gray line). But that choice makes it much much more likely that you'll kill the other guy (red line).
economist.com/interactive/un #TeachEcon
Post
Conversation
The only way to fix something like this is through mandates -- cars/passenger trucks can only go up to a certain weight. And that has about as much chance to pass in the U.S. as single payer healthcare.
This sounds a lot like the problem with the world in general, encapsulated in a single example. The Golden Rule got it a long time ago.
Very interesting. I wonder if we would be better off having all very heavy cars or all very light cars.
There’s also the law of unintended consequences. US tax code incentivizes vehicles heavier than 6000 lbs!
This feels like a great moral hazard problem. All the incentives lead the individual to make the societally-worst decision.
This isn't just a Prisoner's Dilemma under voluntary choice constraints:
Quote
George Singer
@georgewsinger
Replying to @ATabarrok
Regulations and tariffs keep American cars big: vox.com/future-perfect
“But that choice makes it much much more likely that you'll kill the other guy”
Now you know why your car insurance has gone up 50%.
It's important to note that the grey isn't crashes, it's two car crashes.
Single car crashes are much more common. And the most common kind of heavy car, SUVs, struggle to do well in single car crash tests. They're particularly bad with rollovers, which are very dangerous.
fascinating. all collisions between my own car and other cars appear to happen when both cars are exactly 3,700 lbs
This doesnt consider the possibility of dying/getting hurt in a single car accident. Which I think is significantly higher if you’re in a huge vehicle (trucks/suv’s) because they just don’t handle as well.
If everyone drove a 2021 Toyota Camry we’d be on an even playing field at equilibrium. I’m doing my part.
Dumb question: if two 7000lb vehicles are involved in a crash, how do you distinguish between "Own" and "Other"? Or is this just a bad chart?
I'm a safe driver. The people in the "other car" should drive carefully and avoid crashing into me.
That literally makes buying the heavier vehicle the only moral choice.
Thanks, I'll take two.
Plus 'the other guy' could be you... in this escalation of bigger and heavier 'urban assault vehicles'.
It's self-limiting. Everyone will end up with big cars and the problem disappears.
Missing data points here are the frequency and outcome in crashes with large commercial vehicles?
I choose heavier vehicles because they're the only ones capable of doing the work I need my vehicles to do. I'm reasonably confident that the vast majority of the people driving large trucks consider the safety factor a bonus as opposed to a motivation.
Isn't this confounded by the fact that larger cars typically hold more passengers?
Seems like 4,000-5,000 lbs is the sweet spot here where your probability of death is sufficiently decreased without taking too much of a toll on the other vehicle involved.
The difference at the extreme is 3 in 1000. Considering the 1000 is “two car crashes” the difference is infinitesimally small.
Another hyped study meaning nothing.
The high probability of the other person killing you was what started the soccer Mom SUV craze in the first place
Ir anything; this will cause more people to buy SUVs.
Alternatively: Some people choose heavier cars because they require space for more passengers, or more hauling capacity. While others seek larger vehicle for greater visibility, or perceived safety.
I don’t think people buy big cars for safety. The data show beyond 4000lbs the effect is small. They buy big cars because it’s fun and makes them feel powerful.
It also says if you drive a small car you are more likely to die in a crash with another small car.
Obviously, we should encourage the purchase of heavier cars. Why would we want anyone to buy a lighter and therefore deadlier car?
The CAFE regulation are killing Americans. is responsible for many deaths, in order to 'save the planet'.
The conclusion from that article should be:
why the fk does the automobile industry get away with succesfully lobbying for their gas-guzzling SUVs/trucks to get a free pass on environmental/efficiency standards!?
To my understanding this correlation is even stronger with vehicle height. Vehicles with taller grilles (that tend to also be heavier) are exponentially more dangerous while also gradually becoming more popular due to the growth of SUVs and the popularity of "rugged" styling.
It is so infuriating to watch empty pickup trucks drive up and down the road knowing that for almost all of those drivers, a sedan would be plenty large for what they need. The death of the car market in the US has been sad to watch.
Whats really interesting is that although there are declining benefit per pound on survivalbility, lethality increases exponentially
Interesting eye balling own car survival not much apparent advantage between 4000lb and 7000lb; but huge negatives for other car survival across those same weights
I've long wanted some kind of reverse cowcatcher that will lift my nice light car above the enemy's engine block. I get all the benefits of good gas milage and I'm well positioned if I have a head on with a large car.
It would help just about everything if we raised gas taxes. We could reduce taxes in other places to balance it out.
Unfortunately, the weight of most electric vehicles available in the US moves the driver up that red line. An electric SUV can weigh 3 tons or more.
It’s a much better example for teaching unintended consequences:
Do people really take the "weight x safety" consideration as a first order or the "weight x substantial tax benefits for vehicles above 6000"? I would bet on the second one. Maybe a RDD could answer this question.
For this one, I think 3rd party property insurance premiums would help, but less through the price mechanism and more by forcing people to be aware that their choice puts other people’s lives at risk.
But this is still a closed form example. Are not each externalities themselves external in unintended ways? Perhaps a rich description is more valuable than a cheap prescription, because we can choose different paths.
Looks like 4000-4250 would be an optimal limit for people who think they need a heavy vehicle for safety
True, but research on this is at least 20 years old, and perhaps 40. What is new? (One answer: heavy weight of EV batteries. So plug-in hybrids better for traffic deaths). Also continued growing pref for v large vehicles - a 30+ yr trend.
My undergrad major was Public Administration, half business half political science. We had a course in Public Economics. Which covered "externalities" in great detail. I went on to get an MBA and work on Wall Street and nobody at either place knew what an "externality" was.
Isant the issue speed tho? For every 1,000 pounds there's an exponential amount of energy being created. So an average car traveling 60mph has the energy of a cubertruck going 20mph... It's annoying, speed can and should be more effectively controlled by modern tools and policy.
So many comments ignore the systemic effect and assume the "other guy" won't buy a heavier car too. The equilibrium is everybody driving unnecessarily huge cars without any benefit to anyone's safety and with all the costs like increased fuel use & wear on the infrastructure.
Having gone from a Suburban that weighed 3 tonnes to a Hyundai Accent I understand how vulnerable it is and I figure my best defence is not being near big vehicles.
People don’t look for heavier cars bc they are safer. They MIGHT look for safer cars that happen to be heavier.
Watch car ads — ever seen one that says “now heavier with a 6000 lb curb weight!”?
One quick (and, oversimplified) question: if each person buys the heaviest car they can afford (all persons can afford the same set of cars), given the cars owned by all others, then do they all buy the same car?
But this is America. I don't think Americans care about the red line even if you showed them before buying.
I got the heavier car because I have to drive around with 5 car seats and there are no station wagons allowed due to regulations.
I drive a Lexus ES300h (just a fancy Toyota Camry), it's a mid sized sedan. I love this vehicle so much. 44mpg, comfortable. Every feature you could ever want. But always in the back of my mind is the fact that I'm surrounded by small people in giant vehicles who might kill me.
Every time I park my Corolla next to a behemoth SUV, I will think of this study.
I used this in my lecture on institutional economics :) there is even a very nice paper in restud by anderson and auffhammer
Unfortunately, it quickly turns from 'protect myself and my kids' to 'insulate myself from the consequences of driving like a jerk.'
Not sure if it’s a pure externality setup. Isn’t the greater safety of a heavier car a feature that influences the price? I interpret the problem as a social dilemma: NE is the selfish optimum (all with heavier car), but everyone would prefer lighter cars (as they’re cheaper).
Right and in this moral quandary, most people will choose themselves/their families over "the other guy". Regulators have failed us all and instead of browbeating SUV drivers, it's them that should be the focus of our ire.
And a perfect example to teach game theory/ prisoners' dilemma/ Nash equilibrium
Yes justin. I researched vehicle weight exhaustively on that premise alone. You might say I’m a single issue car buyer.
GIF
After I was in a major auto accident my orthopedic surgeon told me to get a bigger vehicle so I'd be the guy who took someone out rather than the guy who gets taken out. So I think this is an explicit thing more than people are implying
this is also where the (inevitable, I think) driverless car AI will end up, I believe - prioritizing the passengers over whoever will be hit
Democracy is a mucho better example, much more intuitive. 51% impose externalities on the other 49%.
I wonder how one-car accidents affect the overall picture of lives saved and lives lost. How many one-car accidents are there compared to 2 of more car accidents?
Graph of basic physics. The real issue is what data was excluded so as to derive these curves. For example, this needs to be controlled for # of passengers/vehicle, & other factors.
In order to do independent analysis, we need the actual data. The conclusion may be more nuanced.
Don’t forget the other negative externalities due to the heavier vehicle, like increased wear on roadways, bridges, and other infrastructure.
The was probably the best article I’ve ever read in The Economist. And when did they start running regressions?
Ironic. Just tweeted about the 8,000 pound Denali Truck. The thing could take out a town. I wish the US invested in mass transit
"Every man for himself", said the elephant as he danced among the chickens.
They are also far more dangerous for pedestrians and cyclists. That's why we need government regulation there.
Is this controlled for type of accident, location, time of day, speed, age of driver, road surface, any of the many variables that could also be involved?
or just wear seatbelts.
80-90% of auto fatalities weren’t wearing seat belts.
Also, what are the whole numbers?
Interesting to see if higher visibility reduces total incidents for some vehicle profiles.
I bought my giant SUV for the tax write-off, but the safety is a nice bonus.
People buy bigger cars for more capacity. Either passenger seats, cargo, or towing.
But you can’t buy a small powerful truck as is common in Japan or Australia because CAFE standards require the longer wider wheelbase.
…
We incentivize cost-shifting onto those who can least afford it in all kinds of deadly ways—the tragedy of the common sense.
This is flawed, larger trucks IE diesels are significantly more likely to be traveling on the highway thus accidents with them have higher fatality rates 
Extremely curious: do those new heavier (hence, more expensive) cars have better anti-crash? And are those crashes the fault of the heavier car?
I'm wondering if the deaths are caused by the other car slamming into the big car and not the other way around.
Shouldn't higher liability insurance premiums reflect the increased risk from heavier cars thus discouraging their purchase?
I’m confused by the graph. If you’re in a heavy car and get hit by a heavy car, which line do you end up on, the red or the black?
Also highlights the importance of regulations. People are selfish and they will do selfish things when given the choice
This is just classic prisoner's dilemma, with an extra push towards the Nash equilibrium coming from suburban mom's in SUVs
That literally wasn't even on my radar when I bought my truck. Towing capability was, though.
I'll make the sensible choice and keep driving my huge land cruiser thanks.
Interesting ethical question ... should you choose a big car that cuts your risk by 30% and increases the risk for the other guy by 500%?
Reading most of the responses here is enough to understand the current state of the union.
Who chooses a heavier vehicle? I’ll take the one that weighs the most please! Vehicles are chosen based on utility, capacity for passengers, hauling, towing, price, style … not weight. By this reasoning no one should buy an EV with its extra 100’s of kg’s in battery weight.
This doesn’t make sense. Because from the ‘other car’s’ perspective he’s driving his own car and should therefore be on gray line.
It’s not terrifying: the base rate is about 1 death per 1000 2-car crashes.
Can't teach negative externalities without mentioning that John Cochrane's existence is a negative externality.
Umm, people do in fact understand that.
And - particularly in your society, where that is the approved ethos - DO NOT CARE!