We are winning the war on automobile deaths | Hacker News Reader

We are winning the war on automobile deaths

evidensity.com

13 pointsbengarvey30 comments

Number of miles driven seems like an important consideration. The following graph is in an interesting (odd?) format, but it looks like the number of miles driven by the average American has dropped every year since 2005:

http://mbostock.github.com/protovis/ex/driving-full.html

On the other hand, the improvements in safety are even more dramatic if you consider the average person is driving about 3000 more miles per year than they were in 1975.

As an aside, the Spark Lines are cute, but if the goal is conveying information it would be nice if they expanded to something more detailed when clicked on, rather than just repeating the same information. I was trying to figure out if reduction in miles driven from 78-80 matched up with the this articles data, but it was hard to tell.

Wow, that graph is awesome. I think it's really interesting how miles driven just gallops forward when fuel is cheap, but falls back only very slowly - or even grows slowly - when fuel is expensive. That looks like a junkie's relationship, for sure.

You seem to imply that accident rates will be proportional to miles driven. I doubt that is true. I would expect the graph of the "chance of an accident per mile" vs "miles driven" to look something like this:

  0
  | *
  | *
  |  *
  |   *
  |      *
  |           *
  |                    *           *          *
  | -  -  -  -  -  -  -  -  -  -  -  -  -  -  -   minimum accident rate
  |
  ---------------------------------------------- 0

I have two reasons:

First, as miles driven increases, driver experience increases. Accidents are much more common among (I'm guessing) 16 year olds with a freshly minted license than a 40 year old.

Second, as miles driven per year increases, the percentage of driving time spent on the freeway increases (due to taking long trips, presumably the major source of milage increase), but accidents are less likely on the freeway. This argument is admittedly much more specious than the first.

Following this reasoning, it is not at all clear that "the improvements in safety are even more dramatic if you consider the average person is driving about 3000 more miles per year than they were in 1975" because those extra miles may not be the most dangerous miles.

First, as miles driven increases, driver experience increases.

I suspect this effect may not be as significant as you think. Certainly, someone who has been driving every day for 20 years will usually be a better driver than someone who has been driving every day for 20 days. I suspect the difference between 2 years and 20 isn't all that dramatic though.

I further suspect that most driving is not practice, by which I mean that drivers are not actively thinking about their driving and how to be better at it. Most of us, once we're reasonably proficient stop actively thinking about the act of driving entirely unless an unusual situation occurs.

If your second reason turns out to be correct, then it also means that a larger percentage of driving than before isn't useful practice for avoiding crashes since highway driving tends to be safe and predictable.

You make good points, but I think that "average miles driven" is likely just the total miles driven divided by population. I'd guess that many (most?) of these new miles are because a greater percentage of the population has a car than in 1975. There may even be relatively more <3000 mile per year drivers now than before. I don't think the linked data is quite enough to strongly support either interpretation. Still, the drop looks more impressive on a per mile basis than on a per person basis.

"As an aside, the Spark Lines are cute, but if the goal is conveying information it would be nice if they expanded to something more detailed when clicked on"

Point taken. The site is a work in progress. I'm working on some cool data visualization tools and trying to use simple charts and graphs to make a case for trends and their causes.

I think miles driven is a very important consideration, but in the short term I had trouble finding data farther back than then 2005 peak mentioned here a few times.

It's not just airbags, a huge part of it is the greater availability of anti-lock breaks and traction/stability control. My brother-in-law is one of the head guys at GM working in this area and he talks about how life saving these technologies are all the time.

There are lots of studies out there on this, this CNN article mentions a study that says that 1/3 of deaths can be prevented by Electronic Stability Control: http://articles.cnn.com/2006-06-09/business/iihs_esc_1_multi...

I have read this also and I think your comment should be more highly ranked. Another link:

http://news.consumerreports.org/cars/2010/04/electronic-stab...

Money quote:

"Consumer Reports believes that electronic stability control is the single most important safety advance since the development of the safety belt."

Thanks for all the comments on this. My next analysis will definitely include stability controls and their adoption rate.

ESC helps in many easy situations but it also allows people to push their driving against the limits of their car. Now then, when ESC eventually fails to correct the vehicle's direction and and grip the crash that follows is usually much worse because there's nothing much that could be done within the laws of physics at that point. I see new cars with ESC at the roadbeds every winter, crashed to something.

Everything certainly comes in good and bad sides but I'm of the opinion that good driving skills and controlled handling of the vehicle in various abnormal situations is worth much more than ESC. I like to feel my car losing its grip and controllability, should that happen, at the point where I can still do something about it.

About 90% percent of driving skills is avoiding unexpected uncontrollable situations and the other 90% is about being able to handle them gracefully.

I grew up in Alaska and saw this sort of thing every fall with four wheel drive. 4wd provides a significant advantage in acceleration on snowy/icy roads, but does not help a car to stop faster or, barring experience with rally driving techniques take a corner at higher speeds. People with 4wd would regularly get overconfident either due to the ease of acceleration or a mistaken belief that their cars could handle slippery corners better than others and end up in the ditch.

I personally hate traction/stability control systems. I feel like the system is fighting my attempts as a driver to get the car to behave in a particular way and really don't want that kind of unpredictability in an emergency. I do have more experience than average controlling cars at the limits of grip than the average driver; I used to autocross pretty regularly.

That said, I'm not opposed to the systems existing or owning cars that have them. I just want the switch to remember my preference instead of having to turn the bloody thing off every time I start the engine.

In my home state of Victoria (Australia) we have gone from 1001 road fatalities in 1970 to just over 300 per year, with a doubling of the state's population to around 4-4.5 million.

I knew someone involved in a sizable study looking at what factors had the biggest effects over time. There were plenty of small factors around certain safety requirements/developments or traffic control, but there were only two big factors: mandatory seat belts and extremely graphic advertisments, aka the TAC ads.

The TAC ads basically showed you what accidents look like in real life, and what you look like after an accident. None of this soft touch crap - the state next door ran an anti-speeding campaign where they had people extending their little finger at speeding drivers (as in 'little dick syndrome'). Absolutely laughable and the folks who are the bulk of the problem - young men - ignore crap like that.

But when you see an advert that has someone's head going through a windshield in slo-mo, or a motorcycle rider coming off his bike crotch-first into an oncoming van, or a hospital scene with a bloke in traction, it helps dissolve the 'it won't happen to me' mindset. Not altogether, but it's no longer conveniently put out of mind.

I'm at work so I can't find some youtubes to demonstrate, but the ads are usually surprisingly 'real'. Needless to say, the ads are usually quite shocking to people exposed to them for the first time.

http://www.youtube.com/watch?v=Z2mf8DtWWd8

Thanks, I didn't realise they had a montage... gets a bit hectic there in the middle.

Another large change I saw in Victoria over those 40 years was a massive change in the public attitude towards the acceptability of drink driving.

They still have some way to go, I think, to catch up with places like Europe where people are shocked if you consider driving after 2 beers.

Victoria seems to have taken all the low-hanging fruit from the road toll, and will have to work hard to get further gains. Many of those remaining 300 deaths are suicides or pedestrians who "made no attempt to avoid the vehicle" which I think means they were drunk and passed out on the road, or something like that.

I didn't read the study I mentioned, just had it related to me by one of the folks involved. You raise a good point, perhaps it may have been hard to determine that very gradual change of opinions of drunk driving as being a large factor - certainly seat belts and the introduction of the TAC ads produced significantly improved road toll numbers in a short period of time, hence are more easily noted.

None of this soft touch crap - the state next door ran an anti-speeding campaign where they had people extending their little finger at speeding drivers (as in 'little dick syndrome'). Absolutely laughable and the folks who are the bulk of the problem - young men - ignore crap like that.

New South Wales (the state next door) has seen a similar reduction in deaths. In about 1970, the death toll was around 1300 per year. It is now down to 367 in the last year, and has been averaging about 400 the last few years. NSW has a population about 30% larger than Victoria.

The ads seen here (NSW) are mostly in in the same graphic vein as those seen in Victoria. Indeed many of them are the same advertisements.

The 'pinky' ad was actually an attempt to reach the young men whom you say would ignore it. It was though that they were actually ignoring the more graphic ads, still thinking that it won't happen to them.

I didn't realise the TAC ads had come through to you blokes - the times I've visited I've only seen ads like the pinky ones. I stand corrected. The pinky ads are still stupid, though - as if the problematic young blokes are going to respond to that :)

I'm thinking its because we didn't "declare war" on traffic deaths. Auto safety has more or less inhabited that strange place where even politicians were inclined to use real science and employ evidence based programs to make driving safer.

If only they'd tried such a sane approach when it came to recreational drugs.

I chose the language in the title intentionally to draw the parallel. I've never heard a single politician in the US even mention it as a problem.

Unfortunately, winning the "war on automobile deaths" does not help those who choose other modes of transportation. More in-car technology creates a more dangerous environment for those outside the car.

Pedestrian deaths have dropped over the time period but not nearly at the same rate as driver deaths. We are seeing more fatalities from distracted and intoxicated drivers.

I think it is about time we stop focusing on safe cars. We need to create roads that are safe for people, not just cars.

Indeed! There's an inverse correlation between automobile safety and pedestrian safety. Cars like wide roads, wide lanes, sparse crossings, road sides with everything removed that you might bump into whereas pedestrians are more safe in pretty much the opposite environment.

I'm JUST starting some data analysis of this in my city so I don't have much factual evidence yet but I heard something like wearing a safety belt doubles your chance of survival.

There are many low budget things that governments can do to help prevent even fender benders (which have an economic cost) and luckily at least in my city it's data driven. Signage, road conditions, weather and light analysis and more play a part.

While here the police don't capture it at the scene of an accident, I bet that pedestrian headphone use causes at least a few accidents. I bike everyday and am constantly stunned that other bicyclists will be wearing headphones will driving around. Kind of need all of your senses out there.

Where did this improvement come from?

If I had to guess, I'd say almost entirely from safety improvements in automobiles. It seems like this perfect little engineering problem (reduce # of deaths/serious injury of people in our cars) which lends itself to continuous improvement over time. Collect and analyze data, develop and apply technology based on results, test improvement effectiveness, iterate forever.

I think there are more game changing safety features just coming into production like:

blind spot monitors http://en.wikipedia.org/wiki/Blind_spot_monitor

Lane departure warning systems. http://en.wikipedia.org/wiki/Lane_departure_warning_system

Precrash systems, like attention assist. http://en.wikipedia.org/wiki/Precrash_system

Great list, thanks.

A lot of this is just because unemployment is 9% and fewer people have jobs to drive to.

Perhaps we should say that people are driving less because there is less commercial incentive to do so. A slower-growing economy has higher unemployment and less need for trucking, etc.

I hope to capture this within a previously mentioned metric, "miles driven."

Reply on news.ycombinator.com