U.S. Traffic Accident Fatalities, 2001-2009

When I was growing up in Australia, the news would frequently report how many road fatalities there had been in Victoria that calendar year. That would normally be done after a report on another road fatality, which in those days there was one of basically every day. They didn’t report the nationwide numbers, but it wasn’t too hard to get the order of magnitude by extrapolating from the Victorian numbers. In other words, I doubt an Australian audience would have been as surprised by the Australian numbers as a US audience is by the US numbers; a similar number is reported frequently on the news.

I wonder how much that level of public knowledge of how deadly the roads are has contributed to Victoria, and Australia’s, incredible decline in road fatalities over the last 40 years. (About 1000/year around 1970 to around 300/year now, despite road usage going up a lot in that time).

Typo–the rate is about 1.5 per 100 million miles.

Fatalities per capita would also be interesting variation on the map. That line down the center of the country’s pretty stark.

In Queensland, people are allowed to place highway crosses where their loved ones died. In the States, that would be a LOT of highway crosses.

It’s common enough in the States on local roads. In fact, I’m writing this comment on my laptop in my local public library just across the street from one commemorating a friend of mine. I’d be able to see it from the window if it wasn’t dark out now (it was two years ago this past Halloween).

Barry, in the UK its full on tribute – particularly if a child was involved. Flowers, I’ve seen bikes, teddy bears, dolls. Quite a recent thing. One day the country seemed to be covered with roadside memorials.

Thanks for presenting this. When we first started visualising this road casualty data we thought we must have got the algorithm wrong because of the huge number of incidents that were showing up. We then confirmed that our rendering was correct and that there were indeed a huge number of crashes and deaths.

I agree that presentation does beg more questions and we will start looking at these, in particular I want to be able to see trends over time for different user groups, different junction, settlements and jurisdictions and for comparisons between different junctions, roads etc.

Are the same roads and junctions the most dangerous every year or is there a lot of randomness in the results?

We have noticed that the percentage of motorcycle fatalities in the USA is much higher than the UK. What is happening to the level of motorbike fatalities over time? Are the numbers increasing or falling and where are these changes taking place?

Looking at the UK data, we noticed that the number of vehicle occupant fatalities in central London has dropped from 17 in 2000 to 1 in 2009 and zero in 2010 but that the number of pedestrian and cyclist fatalities had increased significantly. The London congestion charge was introduced in 2003. Was that a significant influencing factor or not?

Our intention will be to introduce analysis to allow people to explore the trends for themselves in the places that they care about to answer these questions.

I will be following the comments on this blog for any other ideas and questions. Do also look at the UK casualty map for comparison (which has injury data as well):
http://map.itoworld.com/road-casualties-uk

My earlier post was misleading – I was quoting numbers for Victoria, not all of Australia. So Down and Out of Sài Gòn’s calculation isn’t right – and that’s my fault not theirs for being so unclear.

Table 24.25 from this ABS report makes a lot of international comparisons. The US does really badly on deaths per population, but less strikingly badly on deaths per vehicle miles. (Those numbers are from 2007, so a little out of date, especially since fatalities in the US have fallen so much during the recession.)

My understanding is that US fatalities have increased quite significantly in recent decades so it used to be even worse. Not sure where to find the numbers. Kieran?

I was having a conversation with other expats that ended up as a collective rant about the absurdly easy state driving tests, and the generally atrocious standard of driving compared to our home countries. There are other contributory factors: people drive at younger ages, and to older ages. Many of the local stories of road deaths cluster around elderly drivers, or high schoolers late at night, often in rural areas where driving fast on narrow roads, carrying four or five passengers, is the main recreation. In the middle, drink-driving is still not the taboo that it is in most of the developed world.

That’s why David Hembrow’s bicycle advocacy feels idealistic in the US: when you count in drivers who are uncomfortable around bikes, drivers who regard bikes as a personal affront to their masculinity, and drivers who simply don’t see bikes, and factor in road-building decisions that make it hard to find bike-friendly (or bike-legal) routes, the US remains a deeply hostile environment for non-recreational cyclists outside of a few city centres.

A couple of the deaths in my area were pedestrians whose public housing is surrounded on two sides by interstate, and where the access bridge was closed because of concerns about drug-related crime. The only way to get downtown was to take a circuitous route or risk one’s life crossing the interstate.

And as Time Wilkinson points out, no number of lives saved is worth the slightest inconvenience to him.
Ah, if only there were an alternative to trafficking in the most vulgar straw-men imaginable.
More constructively, one could look at Husak’s
“Vehicles and Crashes: Why Is This Moral Issue Overlooked?” 30 Social Theory and Practice (2004), pp.351-370
in which one will find, among other things, an articulation of the difference between bearing the burden of risk of harm done to oneself versus harm done to others.

Back in the ’60s, annual US road deaths were I believe in the 50-55,000 range. We’d hear forecasts and totals for holiday weekends in the 400-500 range. The numbers are down and the rates are ‘way down, because of safer cars (seat belts, airbags, crumple zones, child seats), safer roads, and lower speed limits. People aren’t appalled by today’s numbers because they’re such an improvement.

“people were often surprised to learn just how many road deaths there are in the U.S: about forty thousand per annum (though 2009 saw a very sharp drop, interestingly).”

Perhaps some people who would have committed vehicular suicide in ’09 committed suicide in ’08.

I remember the horror of 101 (California) in the 50’s, both the Bloody Bayshore on the SF peninsula and the stretch from San Jose to South of Gilroy, as being like nothing I’ve seen in maybe forty years. When we drove from Palo Alto to Monterey for a weekend it was almost expected to see a car off the road with a dead body in it.

Also, remember that drunk driving (and I mean bombed, not the current DUI standards) was not just socially acceptable, it was expected, and seat belts met extreme resistance until about 1970.

OP “(though 2009 saw a very sharp drop, interestingly)”

As one who commuted toward NYC at the time, the drop off in traffic after the GFC erupted was immediately noticeable.
I’ve always thought the general level of driving in this NYC metro area to be pretty good (with possible exception of Jersey- “Jersey driver” being a term of abuse in these parts).
This may all just be parochialism/ selective perception, but I tend to agree about driving in the South- they have some pretty bad habits, such as ‘drifting’ lane changes, which we tend to do more sharply up here.
The high toll in northern Australia tracks what has long been known here- the rate per capita in areas like Nevada is far higher than the denser metro areas; miles driven, tedium and falling asleep at the wheel.
I’m surprised by those who say drunk driving is still more tolerated here- evidence?

The numbers don’t surprise me. US numbers have been in the 40K range since I’ve was a kid (about to turn 60). They did go down a bit, primarily because of car design. I once looked up death rates per capita by state, they vary by at least a factor of two. I think when I saw them (decades ago) Wyoming topped the list. Fatalities per vehicle mile are much much higher in the early am hours. This is likely a combination of drunk driving and fatigue.

I’ve done a lot of work on computer codes that car companies use to evaluate safety. A lot of engineering and money goes into getting good scores on the mandated crash tests. Of course the safer you make cars, the more drivers feel they can be more aggressive, so its a lot harder to make progress than if human nature weren’t so perverse.

In 2008 especially high gas prices caused people to slow down, and that saved several thousand lives. Expensive gasoline saves lives. As would lower speed limits that were rigorously enforced. But, politics (and even vehicle marketting) makes that a losing proposition for any politician proposing such.

Re NJ and NYC drivers (they’re interchangeable, in my opinion): aggressive but skilled. That’s the mantra. NY and NJ both light years ahead of Massachusetts drivers — not so much aggressive-but-without-skill (though that’s true) as just batsh*t sloppy and crazy (some Irish Catholic Puritan god knows why). As for foreign standards for gaining a license, I have no idea for most countries. I gather they’re very high in some, like Greece. But there you overcome that problem with a bribe (or refuse to pay up and then don’t get your license, no matter how well you drive).
As for bikes: great where plausible and workable. I don’t see hostility to bikes as a widespread US attitude. It’s just that they’re not on most people’s radar because they’re irrelevant when you’re commuting 20-60 miles each way or live in the northern tier (ice, snow, cold, that stuff). Bikes aren’t the issue. Public transportation is. Only ONE REALLY good public transportation system in the US or Canada: NYC area.

In Australia the legal blood alcohol limit is .05, but more importantly, new drivers have three years of provisional licence when the limit is 0. Any alcohol when you are randomly stopped and you are busted down to Learner. The drunk driving I remember from my youth in the 1980s doesn’t seem to happen with my kids and their friends. By the time they get a full licence, they seem to have internalized the designated driver thing.

Trevor: Given that a large majority of fatal crashes result from some sort of driver error, driver-less cars, if widely adopted, stand to reduce road fatalities enormously.

Reminds me a bit of Piet Hein’s gruk about coffee… but sure, I get your point. Google seem to be enthusiastic about the idea too (one of the ideas they so far haven’t thrown out in their “spring cleaning”). I still think people prefer to feel in control and the feeling of being responsible for your own mistakes, rather than being at the mercy of someone else’s.

36 & 15
I don’t driverless cars will make so much difference unless it is compulsory (and try getting that through Libertarians) – since I suspect that drivers will respond to the predictability of driverless cars by increasing their unpredictability. (For instance reliable maintenance of braking distance will be a lane hopping opportunity.) And there remains the question of liability in the case of accidents involving driverless cars.

I disagree with Matt, no. 3. I rarely drive – I’m a biker – but when I do, I am always amazed at how good people are at driving. They are, after all, speeding along in 3 thousand six hundred pounds of metal in very narrow lanes, they have little elbow room for accident, and yet, like some spontaneous ballet company, they coordinate with amazing grace and agility. This is even more amazing when you consider that each driver seems to consider every other driver an idiot.

Natural history displays a number of marvels – the migrations of birds, whales and butterflies, the strategies of parasites, etc., etc. Surely traffic is one of them.

Benefits of the UK’s advertising industry cluster: really amazingly manipulative and sick road safety campaigns. There was the one with the tombstone popping up in front of the car on a suburban street just too late – inscription TO OUR DARLING SON, in GILL FUCKING SANS type. Or the one with a giant blank billboard with the following message in horrifically crabbed handwriting: “My name is Kate. I am 31 years old. My boyfriend had a couple of drinks before he drove me home. ” Or the one with the guy quietly convulsing in the hospital, or the one with the chap on the bike who looks you in the eye immediately before the crash.

It’s the Los Alamos of propaganda, I tell you. If we ever need to launch a really sinister campaign of genocidal incitement, we’ve got all the makings right here!

Alex—yes, absolutely. I think the most effective such commercial I’ve ever seen is the one from Sussex Safer Roads. It’s unbelievably well-conceived and executed.

or perhaps many in the US do actually need double or triple sized cars

A few do. Modern child seats make children quite a bit safer in a crash than a generation ago where you just put a normal seatbelt on them and hoped, but you can’t sit 3 of them across a bench seat, so if you have 3 children, you actually do need 2 rows of non-front seating if you want the family to be able to go places together in one vehicle.

The thing is, most people can’t afford one car for large family outings *and* 2 more for commuting, so then one of the parents has to commute in the behemoth. (Assuming both parents work and they don’t live near public transit, which in the US is a pretty safe assumption.)

Not saying there aren’t people who buy them unnecessarily, too, but the fact that there’s only one person in it *now* doesn’t mean it wasn’t a good decision to buy such a large vehicle. You have to buy for all trips you’re going to take over a period of years, even though most of them are for fewer people.

Chris:
Up to a point, certainly, but a lot of the giant SUVs don’t actually have that much space inside them given how big they are.

Incidentally another reason for minivans is a lot of people share the school run. Upside of that is less cars on the road; downside of that is bigger cars.

In most cases, spraying bullets at a car behind you with a handheld machine gun is morally unacceptable. An earlier poster remarked about the lazy lane shifts of Southern drivers, however, and I was reminded of the aggressive tailgating endemic to gloomy cities in the north of this country, to which the aforementioned machine-gunning is, too often, the only available remedy.

Jesus, the comments under that last link…they sort of bear out why we need the ads…

This is an important post that underscores the reason why people who work on the social determinants of health have long pointed out that transportation policy is an enormous factor in both overall population health and health inequalities. In the U.S., we decided as a matter of social policy that we would virtually abandon our passenger rail infrastructure in favor of building highways and cars, a decision that obviously had a huge number of political, economic, and social factors influencing it.

That was a choice, and we have reaped the consequences of our utilization of automobiles as a primary form of transit in the U.S. When I speak and write on this, sometimes I note that approximately 17,000 Americans die every year from HIV/AIDS, which means that more than twice as many Americans die from automobile accidents as from HIV/AIDS. And there is little doubt, as Matt points out above, that the riskiest thing most people in the U.S. will ever do is something they do almost every morning (get up and drive their car).

The last point that makes this even more urgent from an ethical perspective is that transportation policy has a dramatic impact on health inequalities (there’s a nice APHA report here:

http://www.apha.org/NR/rdonlyres/43F10382…/TransportationBrief.pdf)

That is, we unquestionably have the power to raise or lower both the absolute number of these deaths and the distribution of them through social policy; better access to public transit especially among vulnerable groups would not only improve overall health but would likely compress health inequalities given that the less affluent are more likely to use public transit in the U.S.

Transportation: another fine example of the seemingly endless string of tragic social policy decisions we seem to excel in making in the U.S.

I was reminded of the aggressive tailgating endemic to gloomy cities in the north of this country, to which the aforementioned machine-gunning is, too often, the only available remedy.

Slowing down until they either get the message, or overtake, can be quite effective.

cian #56

I’m looking for a catchy way to say
“I’ll drive the speed your distance implies”.

I guess something like “Drive closer and I drive SLOWER!”

“The closer you get, the slower we go.”

When I learned to drive in about 1963-4, the number of fatalities was about 50,000 per year. Since the population was then only about half what it is now (maybe a little more, but let’s make this easy) that amounts to an equivalent of over 100,000 deaths per year.

We have Ralph Nader to thank for this, in all likelihood, since drivers continue to be stupid and reckless. To bad Nader didn’t quit then.

@MPA Victoria

I worked on the assembly lines for the suspensions that go into Holden Commodores for a while. I enjoyed the work, assembly line work is not the worst work in the world, and I am glad you’re enjoying the car. What brand is it under?

Three states I know of where people do fail the practical test sometimes: NY, NJ, MA. Also, mandatory liability insurance in those states for registration to be valid. Protection attaches to any driver of the car, though “regular drivers” (the definition of which has become more inclusive, e.g., a grown child who drives the car a bit on occasional visits home) must be listed on the policy, and that drives up the cost of the policy (though you can take them on and off the policy with a phone call — at least, we can). There have been attempts in recent years to get insurance companies to set policy rates by region within a state rather than by the whole state (so that, say, the rest of Massachusetts doesn’t have to subsidize the terrible drivers in and around Boston).

You could fail the test in California (I did once) and I think you still can.

Unlike Greece, I loved my time in Italy but the driving was out of this world — I saw eight accidents in nine months — two of them in front of my apartment on successive days, a record I think would be hard to top.

It’s probably giving credit to one of the few institutions I dislike more than I dislike Ralph Nader, but the insurance industry deserves a lot of credit for sponsoring research and media campaigns that did a lot for traffic safety in the US in the sixties.

On the social background of traffic carelessness, I don’t know how to interpret this, but in the little town of Cottage Grove where my great uncle edited the paper in 1925, there were eight traffic fatalities in one high school class a few years back — the town has maybe 10,000 people in it. That’s brutal.

#54 Dan “In the U.S., we decided as a matter of social policy that we would virtually abandon our passenger rail infrastructure in favor of building highways and cars.”

And very democratically, too: all of the major corporations whose interests were involved apparently got a vote. General Motors, Firestone, and a number of other corporate persons in the 40s decided to buy up and dismantle urban transit systems. So they did. I believe their feeling at the time was that this social policy would better reflect Americans individualism.

In the 80s, I drove an old Chevy with the rear bumpersticker “Please tailgate, I need the money.”

That was a choice, and we have reaped the consequences of our utilization of automobiles as a primary form of transit in the U.S.

Automobiles are THE (not merely “a”) primary form of transportation in virtually every industrialized democracy. They’re just a bit more dominant in the U.S. than in most other countries, because America has more land, higher incomes, and generally newer urban development.

As for safety, the most recent data shows that Amtrak (intercity passenger rail) kills more people per passenger-mile of travel than automobiles. It’s just that most of the people killed by trains are crossing the tracks, not riding on the train.

Hi Emma in Sydney!
All I can say is that I love mine and that the combination of value, size, handling and power is pretty much unmatched on the market. It has been one of the best selling vehicles in Australia for a long time so they must be doing something right.
/ Also I think they are just beautiful cars.

Thing that gave me the most heebie jeebies about driving in the US, the concept of the suicide lane. I’m not sure having an unmarked lane for both directions of traffic to turn (and overtake, at least in practice) from is that good an idea. Governments should pay attention to nicknames.

Comments above prompt me to veer grossly off topic and mention that coincidentally, the Commodores did a track called The Assembly Line.

Not one of the very best on their splendid debut album ‘Machine Gun’ (not when it has ‘Young Girls are my Weakness’, ‘I Feel Sanctified’, ‘The Bump’ and ‘Rapid Fire’ on the other side) though it does have a widely sampled drum break.

And also somebody who can spout crap like this is probably not too up on the ‘data’:
They’re just a bit more dominant in the U.S. than in most other countries, because America has more land, higher incomes, and generally newer urban development.

Cause/effect – even ignoring the deliberate dismantling of infrastructure. No siree, no room for doubt in this man’s mind. He knows the facts and nothing’s going to stop him.

@72 How many of those purported Amtrak deaths involved automobiles coming off worse from collisions with trains on level crossings?

A post with some graphs and links on traffic fatalities in different types of communities in the US. But it’s “respectable” to have your child die in a car wreck, nothing really to be done.

Oops,
that quote was from the post 84 not the post 83 – and forgive the typos (no and between you and me – missing question mark).

Why aren’t people outraged?

Between 1920 and 2000, the rate of fatal automobile accidents per vehicle-mile decreased by a factor of about 17. Except for a pause during the 1960s, progress in reducing fatal accidents has been steady.

http://en.wikipedia.org/wiki/Transportation_safety_in_the_United_States

In my own lifetime, the fatality rate has dropped by a factor of about 5 (from 7 per 100,000,000 miles to 1.5). If people in the U.S. weren’t willing to engage in broad scale social engineering to increase density and use of public transport when the fatality rate was 5, 10, or 15 times higher, it seems extremely unlikely now.

As for international comparisons, if you scroll down to ‘KSI by country’, you find that among developed countries, the U.S. is a bit worse than average for motorway fatality rates per mile, but better than average for non-motorway fatalities:

http://en.wikipedia.org/wiki/Road_traffic_safety

So U.S. drivers, roads, and vehicles are not less safe — people in the U.S. simply drive much more.

87, 89: This paper sums it up nicely for the US on that point. They decompose “fatal crashes/mile driven” into: “all crashes/mile driven” × “crashes with injuries/all crashes” × “fatal crashes/crashes with injuries”. The ratio of rural to urban for these three are =0.67, 1.12, and 2.99 respectively (overall ratio is 2.23). Somewhat more likely to crash per mile driven in cities, slightly less likely to be injured in that crash, and much less likely for those injuries to be fatal.

One last bit of data on decomposing contributors to the fatal accident rate: the second chart down here shows total US fatal crashes by day and time of day (3-hour blocks). Of the 56 time periods, the clear winners are midnight to 3 AM, Sunday and Saturday. Followed by a close grouping of 9PM-midnight Friday, 6PM-9PM Saturday and 9-midnight Saturday. I’m frustrated that I cannot find the data for miles driven broken down that way (I thought I had seen that somewhere), but the midnight to 3s in particular are going to be low on miles driven (even 3-6 AM Saturday and Sunday had more accidents than the mean). In the US people do not live near where they party/drink, nor is there generally any kind of effective transit between the two.

And also somebody who can spout crap like this is probably not too up on the ‘data’:

You’re the one who’s spouting crap.

Highway fatality data here:
http://www.bts.gov/publications/national_transportation_statistics/html/table_01_40.html
http://www.bts.gov/publications/national_transportation_statistics/html/table_02_01.html

Amtrak fatality data here:
http://safetydata.fra.dot.gov/OfficeofSafety/publicsite/Prelim.aspx

As you can see from Table 1-4 in the most recent report, total fatalities by Amtrak in 2009 was 139. This is a fatality rate about of 0.023 per million passenger-miles. That is three times the fatality rate of roads (0.008). In 2009, it appears that Amtrak killed three times as many people per passenger-mile as all passenger and freight vehicles on roads combined.