Ryan Delk wondered over the weekend why modern safety tech, like blind spot detection and lane departure warnings, haven't had any effect on traffic death rates. Here's my guess using an annotated version of his chart:
The safety features that have worked are the ones that keep you safe even if you get into a crash. Shoulder harnesses were made mandatory in the early '70s, and after a few years people started using them and survived more crashes. Air bags became common in the '80s, and a few years later fatalities dropped again. Side air bags became widely used in the 2000s, and a few years later deaths dropped yet again.
Plus there were other safety features introduced along the way, like crumple zones and collapsible steering columns. Those all helped people stay alive.
But the more recent tech doesn't protect you at all. It just helps you drive more safely. As it turns out, though, most Americans either don't use the tech or don't become better drivers because of it. I don't know if I'm typical, but I turn off lane departure warnings because they're annoying and I often don't notice the blind spot warning light until after I've already seen a car in my side mirror.
In the US, at least, it appears there's virtually nothing anyone can do to persuade us to be safer drivers. We chatter, we text, we eat, we cut people off, we fiddle with the radio controls—we'll do practically anything other than keep our eyes consistently on the road at all times. So far there's been no tech capable of changing that.
POSTSCRIPT: Please don't say that deaths are up because of bigger cars and SUVs. The evidence says otherwise, and anyway, common sense tells you that bigger cars might cause more deaths in other cars but would reduce deaths in the big cars themselves. It evens out.
Back in the 80's I read a short article in Psychology Today that outlined a theory that a British Psychologist had come up with. It was called "conservation of misery", which postulated that people would strive to maintain the same level of misery. The psychologist actually used improvements in car safety to illustrate his point, which was that humans would adjust their behavior to nullify the improvements in the safety technology. For example, he noted that as braking systems had become more effective (anti-lock brakes, etc.) drivers started tailgating more.
Portland tries very hard to come up with various bike lane designs to make bicycling safer. Each innovation makes things slightly worse. The only thing that has succeeded is designating certain streets as greenways, which cars cannot use for through driving. You can just drive on them for a block or two to get to your house. That makes things safer for both cyclists and pedestrians. In other words, reducing vehicular traffic is the only sure way to reduce traffic fatalities.
I can't help but find it amusing that Mr. "self driving cars will be here any minute and save us all from ourselves" treats his lane detection software as ineffective and annoying.
You'd think that would be a pretty nifty and rock solid feature on the eve of us all being driven around by Johnny Cabs.
This is pretty much why I don't think self-driving cars will work. I have doubts about all the processing that needs to happen for cars to recognize every possible obstacle or situation and respond properly... but sure, after enough time, humans will resolve that, similarly to how planes can fly themselves just fine. I also have doubts that they'll be terrorist-proof enough: one small hack in the code and cars might be crashing into each other on purpose.
But unlike pilots who eventually are weeded out if they purposefully break rules and terrorists who are heavily screened for at airports, there is no way all the DMVs are going to get serious about weeding out drivers or car owners from modifying their self-driving car to do things like "ignore speed limits" or "run yellow/red lights" or "cut off other cars." And even if they did, there will be a good 10-20 year period in which manually-driven cars following only some of the rules are still mixing with the self-driving cars following all the rules. And just like with Congress, it'll only take a few rulebreakers on the roadway to fck everything up for everyone and create gridlock or pile-ups or self-driving cars that are so cautious that nobody wants to ride in one. (I would totally cut off any car for a parking space that I knew was guaranteed to stop and yield to me... which self-driving cars will be programmed to do to protect their own occupants... and I'm not even that aggressive of a driver.)
Yeah, until human natures changes dramatically in the US, true self-driving cars are going to either be slow and annoying to ride in (as everyone else takes advantage of them) or be glitchy and dangerous to ride in (as we see with Teslas today). If even a tech-optimist like Kevin can't be persuaded that The Car Knows Best So Do What It Says at every moment, what hope do we have to convince tens of millions of other drivers to do so?
I sure hope they do work (and I think that you are being too pessimistic). It would be a boon to old people who cannot (or at least should not) be driving as well as parents of teens without licenses who want to send their kids to soccer practice or whatever.
Why do you think automated cars have more trouble with bad drivers than other bad drivers?
So collisions between fully-loaded semis must have very "even" death rates.
WTF?
I’m also unconvinced about the common sense. Think about 3 worlds: all tiny cars, all big cars, and a 50/50 mix. I argue that the fewest deaths would occur in tiny car world, the 2nd fewest in big car world, and the highest in mixed world. Even before we consider collisions between cars and pedestrians or cars and bikes.
Small cars have smaller crumple zones.
In general, identical collisions between large cars are more survivable than identical collisions between small cars.
That's not quite true.
And it ignores vulnerable road users.
...which seems on-point for the guy arguing that the murderer motorist who said he wished to kill some protestors and then killed someone who confronted him for driving his car into a crowded crosswalk.
The big car world would kill the most pedestrians and cyclists, tho.
When you have a point to make and don't want to do any research, just hand-wave it away with "it all evens out."
Also, in southern California, I don't think cyclists or pedestrians exist. If they do, they must not matter.
+1
Yes, the postcript portion is a bit comical.
On the contrary, the lane-departure warning system on my car has trained me to always use the turn signal to change lanes, with the result that the LDW very rarely triggers.
You may be in the minority on this one. If anything it seems that drivers today are less likely to signal lane changes than they used to be. If one accounts for biases in being more aggravated by what is happening now, it is still likely that things have not gotten better in this regard, even if a driver like yourself is being better.
I drive a Honda that has a passenger side mirror camera which is far superior to the lane-departure warning we have on our new car. For some bizarre reason Honda stopped making those cameras to be more “digital” with the lane-departure warning (so I was told by the dealer) .No other manufacturer ever made side camera but it’s is nowhere near as effective as the side camera view. What I really like is that when driving on multi-lane roads you have the option of keeping the camera on at all times, not just when using the blinker. It gives me a great view of traffic in the lanes to the right of me including any in the blind spot.
Don't you have to allow for the slow replacement of older cars with newer cars on the highway? Maybe the year when the mean car would have the feature?
Hence the delays in the graphs.
It may be that it will take a while for driving habits to change to make best use of the new technologies. I've recently bought a car with little light-up arrows on my side mirrors to show if there's someone in my blind spot. I've found myself a couple of times actually checking for the light rather than (in addition to) just looking in the mirror before changing lanes. It's not a habit yet, but in time it might be, and people who have that tech from the start of their driving life might internalize it, but the effects won't show up until there are more of them.
On my car, I'm pretty sure the way this works is, if there is a car in the blind spot, and I change the turn signal to indicate I'm going to change lanes, the system beeps at me. I like that warning.
Same here. If I use the turn signal OR start to move into the lane, I get beeped at.
The Insurance Institute for Highway Safety disagrees that driver assistance doesn't help. They track police reports and insurance claims as a function of assistance features (see https://www.iihs.org/media/290e24fd-a8ab-4f07-9d92-737b909a4b5e/HvQHjw/Topics/ADVANCED%20DRIVER%20ASSISTANCE/IIHS-HLDI-CA-benefits.pdf) and have found that emergency braking reduces front-to-rear crashes with injuries by 56%, just to pick one stat. Blind spot detection appears to help the least, reducing lane-change-crashes-with-injuries by 23%. These seem huge! There was a big dip in 2010 that these may be responsible for. Looking at just fatalities might not be showing the whole benefit.
If somebody doesn't die, then obviously it doesn't matter. We saw this with covid infections too - it's no worse than the flu! (please ignore the millions reporting long covid symptoms) - and we see it here with car crash statistics. As long as the overall number of deaths are going down, the Kevin Drums of the world will be happy to ignore all the non-death impacts like increases in the number of people paralyzed or permanently disfigured or whatever else may be going on.
There's a sharp upturn in 2020, when Covid got underway. Coincidence? I noted that, while streets were largely empty since folks weren't going to work or school, some drivers would take advantage of it by using the major streets like drag race tracks. Maybe a lot of people formed bad habits?
That is still going in my area which makes me think the increase may be due to those drivers. A lot of the people racing are driving very loud muscle cars which is a violation of the noise ordinance — which is not enforced.
There was a Canadian study in Toronto showing that getting a COVID vaccination lowered one's risk of hospital admission for a traffic accident. This result was cross sex, cross age and so on. I seriously doubt that a COVID vaccine improves one's resilience in an automotive collision. The article delicately hinted that the kind of person willing to risk COVID is also more likely to drive like an asshole.
I'm not exactly sure how this explains persistently higher traffic death rates. Could COVID have unleashed the asshole on the road, first by cutting traffic, then by lowering the value people place on their lives?
Reference: https://www.amjmed.com/article/S0002-9343(22)00822-1/fulltext
So, pedestrians don't matter to Kevin, then. Good to know if I'm ever in Orange County and want to take a walk. He may be an examplar of local attitudes.
Far from being a local attitude, it's basically the driving and road design philosophy of the entire North American development/construction paradigm.
SUV's were having issues with rolling over (top heavy). Granted that's only a small percent of accidents and cars have "roll cages".
Of course, if you're a pedestrian....
The problem with the upsizing of American cars is that once one person does it, everyone feels like they have too. Tall cars block your vision, well you need a tall car. Afraid of a bigger car hitting you, well you need a bigger car. We're all in an arms race.
Taller cars are also a lot easier for people with arthritis and other mobility problems to get in and out of.
I think it is wrong to assume that because a piece of technology tries to be helpful that it is helpful.
Lane departure alerts become nonsensical when you live where there is snow. As the roads get narrower the "feature" beeps non-stop but there is nowhere else to drive but over the line. It becomes dangerous when combined with Lane Departure "assist" features, which try to return you to the proper lane (steer you into a snowbank). I have also had to fight against the assist feature because it was trying to steer into a pedestrian (I had drifted over the line going around them!).
phones
Gonna say the same thing.
But I also wonder about whether screen controls (rather than nobs or buttons) might also be at play here. While I might just be an old fart, I swear that changing radio channels or adjusting the AC with my damn little screen rather than just twisting the dial will kill me some day...
Speaking of safety tech, didn't some bro promise people to have completely autonomous vehicles by now?
There are, in some cities. But we don't seem to have the will to make the roads or create standards for them that can be proven.
I wouldn't consider Waymo to be completely autonomous. They have people watching the vehicle remotely, including using cameras inside the vehicle, the entire ride.
It may be difficult for the tech to make up for the greater level of distraction today: cell phones and large screens in cars are just going to be attention-grabbing. There's no way around that.
It's crashes with other motor vehicles (and pedestrian deaths) that are really up: https://injuryfacts.nsc.org/motor-vehicle/historical-fatality-trends/deaths-by-type-of-incident/
That suggests that there could be something to the whole "bigger cars and SUVs" issue. I had thought that perhaps the higher risk of rollover with SUVs was an issue (it was in the late 90s), but non-collision incidents are actually down over the last decade. Drivers are killing each other more often and killing pedestrians.
If this were a Drum post there'd be some snark about people not looking up the numbers. 😉
'Large screens' are just dash boards. There is no study saying such cars get into more accidents.
One reason the death rate has not gone down recently in the US unlike almost every other comparable country is that drivers are getting worse in the US because there seems to be way less enforcement of traffic laws in the US. I live in Canada now and lived in the UK, Malaysia and France. US traffic laws enforcement is more like Malaysian than any of the other counties.
"...because there seems to be way less enforcement of traffic laws in the US. .."
Also enforcement has decreased in recent years leading to higher average speeds on the interstates for example.
The interstates are the safest roads we have. It's the higher speeds non-freeway arterial roads, the "stroads" connecting suburban developments and poorly designed country roads where we see the most fatal accidents.
1000% agree. As an Australian (who as a country have a similar high level reliance on cars to the US and a similar love affair with overly large cars) who has driven in the US a fair bit it is very clear that speed limits are a suggestion and not anything like a limit. If the US had half the enforcement and therefore compliance that Australia does then I suspect the death toll on its roads would be a lot closer to Australia’s (which is a third of the US’).
I have read that nothing has had the effect of decreasing injuries as much as the seat belt. After shoulder belts, injuries dropped so dramatically that subsequent improvements have only chipped away at the remaining cases, partly by turning once-fatal collisions into injury crashes.
It could be that driving is so interently dangerous that once you have eliminated the worst outcome - being thrown clear in a head-on or rollover, further reducing harm involves preserving humans strapped into a protective enclosure.
ESC has to have prevented quite a few injuries and deaths, on top of ABS.
I wish we could go to rail, and kiss these jalopies goodbye.
I haven't seen data to address Drum's last "Update" on how things evening out in collisions between cars. But I have seen quite a bit of evidence that larger cars are very much more hazardous to not cars like pedestrians and cyclists.
And maybe OT, but it's one of my many frustrations with the U.S. that the EV market here will, instead of making small EVs will, of course be making huge EVs. Electrifying the fleet seemed like a good time to start making smaller vehicles, but this is the U.S. So, we'll have 10000 lbs huge electric vehicles. I just hope that my little Mazda S3 never has an encounter with a 9000 lbs F-150 Lightning.
A 9000 lb Lightning has 2500 lbs of cargo, so... it's no different than hitting any other 9000 lb gross weight truck.
" Please don't say that deaths are up because of bigger cars and SUVs. The evidence says otherwise, and anyway, common sense tells you that bigger cars might cause more deaths in other cars but would reduce deaths in the big cars themselves. It evens out."
I thought studies showed exactly the opposite. The street engineers also tell me that bigger cars not only are more dangerous IF there is an accident, but also more dangerous because they LEAD TO MORE ACCIDENTS because, well, they're BIGGER and so harder to stop and turn, as well as just taking up more space.
Any evidence that bigger cars take longer to stop?
Remember - they have more momentum but they also have more frictional force to stop them.
Ugh, there's a limit to how much friction you can put to the road. Tires have a limit.
How do they have more frictional forces? Are the breaking pads relatively bigger in a proportional way?
The safer a driver feels, the more likely they are to drive unsafely or distracted.
Also, the data very much supports the point about the impact of vehicle types.
According to the Department of Transportation, in 1994 the percentage of fatality crashes by vehicle were these:
Passenger vehicles (Sedans and similar): 55.1%
Light Trucks (SUV, pickups): 29.8%
By 2000 the numbers were now:
48.3% (PV)
35.6% (LT)
By 2010
38.7% (PV)
39% (LT)
By 2021
34.4% (PV)
41.9% (LT)
Considering that over 60% of all fatal crashes involve "Going straight", meaning head on collisions, it makes a lot of sense that as vehicles get more massive, improvements in safety features would eventually not be able to keep up.
You need to normalize for percent of each driven on the roads. Most cars on the roads are considered light trucks.
It’s all about the enforcement of road rules and specifically speed limits. Australia, like the US, has spread out cities that are reliant on cars and a love of big cars (not as much as the US but closer than Europe). However Australia’s death tolls is 4.3 per 100,000 while the US’ is 12.9. Much closer to India’s (with its famously lax road rules) 15.6.
The difference? Enforcement of and therefore compliance with speed limits and other road rules. As an Australian who has driven a fair bit in the US it is clear speed limits are a suggestion and not a rule. Just look at proliferation of motor vehicle accident lawyer billboards in Cali (something I’ve never seen in Australia) to see the consequences of this. If the US wants road fatality tolls in line with other developed countries then it needs to employ developed world enforcement otherwise it will continue to achieve developing world outcomes.
Yes. We're all driving much much faster
I may be mistaken, but i thought car fatalities have continued to decline in the EU area. There are also wide divergences in death rates between US states.
Instead of shrugging our shoulders and simply assuming that nothing can be done we could consider what is different between the areas with higher deaths and lower deaths.
Assuming that Americans cannot be 'persuaded' to be safer drivers while also doing literally nothing to create safer driving conditions is a bit, what's the word.....dumb?
Massachusetts had a much lower automotive death rate. A lot of this is because the roads are so crappy, laid out by cows, that it's hard to build up a lot of speed. There are a few heavily used faster roads and some people try to drive like the getaway van in the movie The Town. Interestingly, the opening bank robbery in Harvard Square was based on a real incident. In the movie, the robbers got away. In real life, it was Harvard Square. They gave up their getaway car and tried to run for it. They were all caught.
Traffic laws, highway improvements, and vehicle technology are fundamentally irrelevant.
1980: 51,000 total traffic fatalities (25,000 involved drunk drivers)
2021: 43,000 total traffic fatalities (13,000 involved drunk drivers)
All the improvement in traffic deaths came from fewer drunk drivers causing fatal accidents.
There's a big logic fail here.
The evidence does *not* say that larger SUVs are not related to the increase in fatalities, tho, Kevin.
Cars today easily cruise at 80 mph while keeping the occupants in quiet comfort, while 80 mph in a 1988 LeSabre would've been a pants wetting experience. What's the data for average highway speeds over the years and how does that correlate?
The change in vehicle profiles over the last 20 years is responsible for an increase in fatalities.
Insurance Institute for Highway Safety (IIHS)
Whatever their nose shape, pickups, SUVs and vans with a hood height greater than 40 inches are about 45 percent more likely to cause fatalities in pedestrian crashes than cars and other vehicles with a hood height of 30 inches or less and a sloping profile, an IIHS study of nearly 18,000 pedestrian crashes found.
https://www.iihs.org/news/detail/vehicles-with-higher-more-vertical-front-ends-pose-greater-risk-to-pedestrians
I use the WIPS system-- "Wife In Passenger Seat"-- to regulate my speed and lane control. I turn the lane warnings off because every time I shave a corner and it goes off, I get a protest from my right. Ditto when I put on my turn signal early to warn following cars that I am changing lanes after this guy passes me-- "Why is the car beeping?". I have used the lane-keeping assist on the cruise control on the highway, but I LIKE to drive. I can't understand these folks who buy a sporty car just so they can sit back and let it drive itself. Those Ford (?) truck commercials: sure, I'm comfortable towing a trailer on a mountain road while my hands are in my lap. JFC.
Rear-view cameras, though: love 'em. I will park nose-in in a lot rather than pull through to be nose-out: the camera lets me see traffic behind me much better than sticking my hood out while trying to peer over the SUVs and pickups. And I just did a (rare) parallel park, and could zip in, in one smooth move. Awesome.
There was a profile of some wealthy guy back in the 1960s. He had a driver for his fancy sports car. Someone remarked that this was like hiring someone to have sex with your mistress.
We have data on traffic deaths and a post that discusses exclusively the safety of people inside a car. Leaving out pedestrians and bicyclists (not sure about motor cyclists).
These groups have "contributed" much to the number of victims in recent years. I'd give this post a grade F.