Car Fatalities in America
One story you’re not going to see leading tomorrow’s newspaper is “97 dead in fatal car accidents.” And yet in 2007, this country saw 37,284 people die in car wrecks. That averages out to 97 per day—much more than the seven people whose death in yesterday’s Metro crash has acquired so much coverage today. Obviously in part that’s because driving is much more popular than transit. Still, according to the Census Bureau 87.7 percent of people get to work either by driving alone or in car pools, while 4.7 percent take transit. That’s about 18 times more driving than transit usage. By contrast, 14 times more people die in car wrecks on an average day than died on the rare day that anyone died in a train crash. On a typical day, of course, the United States has zero train-related fatalities.
Long story short, investments in mass transit would have substantial public health benefits. And, indeed, since car wrecks disproportionately affect teenagers and young adults the impact in QALYs of even moderate reductions in automobile usage would be enormous. The good news about this, however, is that the death rate per 100 million VMT has been declining in recent years:
This is rarely discussed, but the consequence is that 1,000 fewer people died in car wrecks in 2007 than died in 1994 even though total vehicle miles traveled increased from 2,358 billion to 3,030 billion. That’s a huge gain for the country.
45 Responses to “Car Fatalities in America”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
June 23rd, 2009 at 4:47 pm
I think that the Aztec empire at its height sacrificed about 30,000 a year. They obviously can’t hold a candle to us.
June 23rd, 2009 at 4:51 pm
Long story short, investments in mass transit would have substantial public health benefits.
Although I think you’re probably right, you are assuming that the fatality rate on trains would stay constant if the amount of transit were drastically increased.
Also, keep in mind that most driving fatalities are due either to alcohol or to failure to buckle up. If you don’t drink and drive, and always buckle up, your chances of dying in a car crash are much lower. Whereas you can’t do much to make your train ride safer.
June 23rd, 2009 at 4:54 pm
This is rarely discussed, but the consequence is that 1,000 fewer people died in car wrecks in 2007 than died in 1994 even though total vehicle miles traveled increased from 2,358 billion to 3,030 billion. That’s a huge gain for the country.
That’s 3 people a day. How many deaths are attributable to the increased air pollution from the increased VMT/yr.?
June 23rd, 2009 at 4:57 pm
Its true that not drinking and buckling up are important, but you can be killed by other people not drinking. Long story short we should increase taxes on alcohol and use the money on better transit systems.
June 23rd, 2009 at 4:58 pm
Really?
According to this page, total traffic fatalities were 1000 higher in 2007 than in 1994. The number of auto passenger fatalities did drop (by 1500) but this was easily balanced out by a more than doubling in the number of motorcycle fatalities.
Not to take away from the larger point — the highways are clearly safer now than 15 years ago — but this seemed like an odd numerical mismatch in your writeup.
Also, more good news: collateral damage (nonmotorist fatalities) also dropped significantly.
June 23rd, 2009 at 4:59 pm
Still, according to the Census Bureau 87.7 percent of people get to work either by driving alone or in car pools, while 4.7 percent take transit. That’s about 18 times more driving than transit usage. By contrast, 14 times more people die in car wrecks on an average day than died on the rare day that anyone died in a train crash. On a typical day, of course, the United States has zero train-related fatalities.
I would think you would want to look at it on the “vehicle miles traveled” basis you use below, since people could make longer commutes in cars than on subways, etc.
June 23rd, 2009 at 5:01 pm
I noted this at the tail end of the earlier train thread, but I’ll repeat it here.
The News Media is doing a really shitty job covering this story — because they don’t show the APPROACH to the accident. Hence, there’s is no way to tell if the operator should have been able to throw on the brakes. It looks to me like the fence between the tracks could have obscured the parked train.
This is important because it raises the issue of whether Metro management are operating the trains at UNSAFE speeds during rush hour –to hand a lack of capacity problem — and depending upon a faulty computer system to avoid collisons.
By UNSAFE conditions, I’m referring to speeds too fast to allow a normal human operator time to throw on the brakes.
A good driver adjusts speeds depending upon how far ahead he can see — but some news reports indicate that the computer system takes control at rush hour and governs the speed. Operators supposedly can override, but I wonder what the instructions from management have been re punishment for doing so.
Maybe the operator wasn’t watching because she was new and had been intimidated into NOT slowing the train down in areas of limited view. To be a “Team Player” and ensure the trains run on time.
June 23rd, 2009 at 5:02 pm
You don’t have to go to the trouble of mounting a coverup when the news reporters are too fucking stupid to even realize a problem exists.
June 23rd, 2009 at 5:09 pm
37,284 people die in car wrecks. That averages out to 97 per day
37,284/365 = 102 per day
June 23rd, 2009 at 5:12 pm
Although I think you’re probably right, you are assuming that the fatality rate on trains would stay constant if the amount of transit were drastically increased.
Not constant, just low enough for the purposes of Matt’s argument. Which is a very safe assumption.
Also, keep in mind that most driving fatalities are due either to alcohol or to failure to buckle up. If you don’t drink and drive, and always buckle up, your chances of dying in a car crash are much lower.
But not lower than your chances of dying in a train crash. Crashes involving alcohol-impairment account for around 1/3 of fatalities, but that includes fatalities where the person killed wasn’t the one impaired (drunk drivers don’t just kill themselves). Meanwhile, around 60% of fatalities involve a person not wearing a seatbelt, but around 20% of people don’t wear seatbelts anyway.
But let’s be very generous and assume you could prevent (.333) and a further (.6) fatalities by not driving drunk and wearing your seatbelt. You’d cut your fatality rate to about 1/4, which is good . . . but not nearly enough to make driving less dangerous than trains. For that purpose you would need to cut it more like 1/14, and that just isn’t possible with the stuff that is solely in your control.
June 23rd, 2009 at 5:13 pm
This analysis assumes that the car trips that would be substituted by transit are all in the average range of danger.
But this is belied by MY’s statement that “car wrecks disproportionately affect teenagers and young adults.” Most car wrecks aren’t rush hour commutes that would go away if we has a better transit system. And even if it is teenagers commuting to school, there is already a subsidized transit option (school buses) that they are explicitly rejecting.
June 23rd, 2009 at 5:23 pm
Rubble,
You must be new here.
Matt Yglesias desperately needs an editor. He constantly uses the wrong word by accident.
This habit is most irritating when he does something like this, and used “fewer” instead of “more.” Sometimes he drops a “not,” which is always fun.
It’s pretty clear from the context – providing the total vehicle miles travelled after an “even though” – that he meant to make the same point as yours.
June 23rd, 2009 at 5:24 pm
IIRC, over half of drunk-driving fatalities are people hit by, or riding with, a drunk driver.
June 23rd, 2009 at 5:44 pm
This point was made by several others in the earlier thread, but people are interested in spectacular accidents. The Minnesota road bridge collapse got plenty of media coverage. The 18-wheeler swerving to avoid a car and going off the Chesapeake Bay bridge also garnered a lot of attention.
Metrorail “got lucky” as coverage was minimal of their previous fatal Metrorail passenger accident in 1982. As I recall, the accident happened about the same time as the Air Florida plane crashed into the 14th Street bridge.
Matt’s other point is interesting about the decline in fatalities for VMT. I wonder how much new safety features/added weight to vehicles has decreased fuel economy…?
June 23rd, 2009 at 5:45 pm
How many deaths are attributable to the increased air pollution from the increased VMT/yr.?
I can’t speak for elsewhere, but at least in Southern California, due to increasingly stringent standards, air quality has gotten much, much better over the past three decades.
http://www.aqmd.gov/smog/o3trend.html
June 23rd, 2009 at 5:52 pm
Matt’s other point is interesting about the decline in fatalities for VMT. I wonder how much new safety features/added weight to vehicles has decreased fuel economy…?
I’ll respond to my own post… meant how much higher would fuel economy be w/o the added weight of new safety features?
June 23rd, 2009 at 5:52 pm
@Adam Villani
Sure, over the past 3 decades air quality has improved in SoCal and elsewhere. The time span Yglesias is talking about is 1994-2007.
June 23rd, 2009 at 6:05 pm
The Transportation Research Board’s Executive Committee periodically identifies a set of critical issues in transportation to focus attention on their likely impact on the nation’s economy and quality of life.
“The 2009 Critical Issues update elevates the importance of energy and environmental issues to reflect the prominence that these topics have gained in national debates about energy security and climate change. Greater emphasis also is given to the issues of the condition and financing of infrastructure, to help policy makers prepare for the reauthorization of federal surface transportation programs that expire in 2009.”
for a copy of report:
/CriticalIssues09.pdf
June 23rd, 2009 at 6:31 pm
most auto accidents aren’t caused by drinking, despite what MADD’s ’statistics’ might have led you to believe. Cars are just inherently dangerous. Thats sort of teh point; careening around in a metal fram is a big thrill.
June 23rd, 2009 at 6:48 pm
And most alcohol related fatalities that are actually caused by alcohol involved people significantly impaired, not just legally over the limit. alcohol’s impairing effects (including deciding if you’re ok to drive) follow more of a exponential curve than a linear one. So ‘don’t drive if you had a drink’ is going to do very little for your chance of dying.
June 23rd, 2009 at 7:01 pm
DTW:
But let’s be very generous and assume you could prevent (.333) and a further (.6) fatalities by not driving drunk and wearing your seatbelt. You’d cut your fatality rate to about 1/4, which is good . . . but not nearly enough to make driving less dangerous than trains. For that purpose you would need to cut it more like 1/14, and that just isn’t possible with the stuff that is solely in your control.
Where does that 1/14 number come from? A random number picked out of Matt’s post and applied to the wrong thing?
So, say we have 5326 times as much death from car accidents as train accidents (i.e. one DC Metro type train accident per year). Now scale that down by the driving/riding ratio of 18, and you have 295 times as much death from cars as from trains. If you wear your seat belt, it’s about 99. If you don’t drink, it’s probably lower.
Now assume that there’s a lot more trains on the tracks and a lot fewer cars on the road. That number will go way down. Data from the UK indicates that rail is about 1/6 as dangerous as driving.
Which means Matt’s point is still right. It’s just not AS right as it initially appears.
June 23rd, 2009 at 7:33 pm
> Although I think you’re probably right, you are
> assuming that the fatality rate on trains would stay
> constant if the amount of transit were drastically
> increased.
Personally I am a big fan of more transit, but MY should to to the library and scan through some issues of “Trains” and similar publications from the 1920s, 30s, and 40s. Far more passenger-rail-miles traveled then and far more accidents. In one sense technology was less advanced than today (e.g. no on-line sonar scanning of rails) but on the other hand a _lot_ more experienced railroad engineers, maintenance foremen, and maintenance workers keeping the plant up than today.
Cranky
June 23rd, 2009 at 8:11 pm
“You’d cut your fatality rate to about 1/4, which is good . . . but not nearly enough to make driving less dangerous than trains.”
But again, we are talking apples and oranges in terms of miles traveled. Halving the number if people who drive and putting them in transit does more than change the head count. People who drive to work often drive really, really far. So while you double the number of people, you more than double their exposure to the dangers of travel. Add to this that cramming all those people onto the trains will probably lead to an increase in capacity. But probaly not a doubling. I preseume the extra wear-and-tear on trains, plus pressures to stick to schedules, would have to increase accidents somewhat.
Finally, getting a bunch of people to use transit does not necessarily eliminate their exposure to the dangers of driving. Most people I know who use transit drive to the lot and get on the train. I presume this trend would be even more pronounced if we halved the number of drivers.
June 23rd, 2009 at 8:35 pm
I’d die of boredom or a premature heart attack if I had to take a train to work every day instead of riding my motorcycle.
June 23rd, 2009 at 8:39 pm
If the highways were now commercial ventures, as once in our history they were, and upward of 40,000 people were killed on them annually, you can bet your bottom dollar that Ted Kennedy and his ilk would be holding Senate hearings on the matter. Blamed would be “capitalism,” “markets,” “greed,” i.e., the usual suspects. But it is the public authorities who are responsible for this slaughter of the innocents.
Most evidence suggest half or fewer people would die if government ceased to own highways, but don’t expect MattY to bring that up.
June 23rd, 2009 at 8:46 pm
Where does that 1/14 number come from?
From this oft-cited Dartmouth “Chance News” study, finding a .000000088 probability of death per 100 miles travelled in trains, and .00000117 probability of death per 100 miles travelled in cars, which is a ratio of 13.3 (I previously screwed up the math in my head so misrounded):
Chance News
By the way, countries like the UK can have very different fatality statistics for modes of transport, so I wouldn’t rely on them for U.S. cases.
Now assume that there’s a lot more trains on the tracks and a lot fewer cars on the road. That number will go way down.
I don’t think that is a valid assumption. Congestion can increase accident rates, but it is far from the only factor. Plus, we don’t know that congestion would be increased for trains in Matt’s hypothetical (although I do assume he would plan for a decrease in highway congestion). So, I’m not sure the ratio would change that much.
June 23rd, 2009 at 8:47 pm
“Most evidence suggest half or fewer people would die if government ceased to own highways…”
That I’d like to see. Hit me!
June 23rd, 2009 at 8:53 pm
But again, we are talking apples and oranges in terms of miles traveled.
As I just noted, these were probabilities of death normalized over miles traveled.
Add to this that cramming all those people onto the trains will probably lead to an increase in capacity. But probaly not a doubling. I preseume the extra wear-and-tear on trains, plus pressures to stick to schedules, would have to increase accidents somewhat.
This is Matt’s hypothetical, so we don’t really know what sort of funding levels, loads, safety measures, and so on we are talking about. Depending on the assumptions, the death probability per 100 miles travelled could actually be lowered, not increased.
Finally, getting a bunch of people to use transit does not necessarily eliminate their exposure to the dangers of driving. Most people I know who use transit drive to the lot and get on the train.
Sure, so really we need to think in terms of the net passenger-miles effects, with the likelihood that train miles won’t swap out for car miles on quite a 1:1 basis. But again, with something like a 13:1 death probability ratio to start with, you don’t need anything like a 1:1 swap ratio to come out way ahead.
June 23rd, 2009 at 8:56 pm
Most evidence suggest half or fewer people would die if government ceased to own highways, but don’t expect MattY to bring that up.
Our resident libertarian-types have brought this claim up a few times now. My guess is that they aren’t controlling for other factors, but in any event it is long past time to actually cite or link something in support of this contention if people want it to be credited.
June 23rd, 2009 at 10:55 pm
Also, keep in mind that most driving fatalities are due either to alcohol or to failure to buckle up. If you don’t drink and drive, and always buckle up, your chances of dying in a car crash are much lower. Whereas you can’t do much to make your train ride safer.
But if you drink, like I do, and you can take a train home rather than driving, that would dramatically reduce the risk not just to yourself, but to others on the road as well.
Good public transportation has definitely saved me jail time, and possibly my life as well. That’s another reason why it’s such a worthwhile thing to invest in.
June 23rd, 2009 at 10:57 pm
(BTW, Nathan @ 25 is hilarious. Is that a real post or a parody troll?)
June 23rd, 2009 at 11:11 pm
Most evidence suggest half or fewer people would die if government ceased to own highways, but don’t expect MattY to bring that up.
Government ownership of highways is reasonably common in Europe, and has been experimented in places like Indiana and Illinois.
There is no evidence whatsoever to suggest that fewer people die on these roads.
June 24th, 2009 at 12:45 am
“There is no evidence whatsoever to suggest that fewer people die on these roads.”
In Brazil fewer people die on these privatized roads.
June 24th, 2009 at 3:28 am
“Most evidence suggest half or fewer people would die if government ceased to own highways”
The private roads are safer because they are newer and nobody drives on them. They are too expensive. Turns out, private industry isn’t better for some things.
June 24th, 2009 at 4:01 am
Teenagers and young adults rarely get killed in wrecks while commuting to work. So increased investment in transit would make little improvement there. The effect you describe is far less than you claim.
But it’s good that your record for producing stupid, wrong and ill-researched blog postings remains unsullied.
June 24th, 2009 at 9:12 am
Safety features reduce pollution.
In 2003, the average new car weighed 4021 pounds. How much the safety features weigh is not as easy to determine. Is the chassis a safety feature? The bumper? Breaklights, tail lights and headlights? If you just mean things like airbags, stabilization control or energy absorbing bumpers (minus regular bumpers), you might get a 1% reduction in mileage. That 1% reduction in mileage doesn’t come close to making up for the pollution resulting from building new cars to replace those wrecked in the excess accidents.
Even the deaths cause pollution. Societies tend toward population replacement. Where people die younger, birthrates are higher, generally. People who die shortly after maturity have a disproportionately high environmental impact compared to their contribution to society. Each person contributes thousands of diapers to landfills, puts pollution in the air for heating and airconditioning, requires food to be grown, and trucked in to the local supermarket etc. If they die at 18, they’ve inflicted all of that damage on the environment, and contributed almost nothing to society.
June 24th, 2009 at 2:43 pm
Are folks seriously debating which is safer, cars or trains?
All I know is that nobody I know has ever been killed in a train wreck.
And everybody here is talking about fatalities, but I know a number of people who are physically and mentally crippled forever because of injuries sustained in car accidents.
We like to think that we are in full control of what happens on the road, but we aren’t. Don’t kid yourself.
June 24th, 2009 at 3:52 pm
Sure, over the past 3 decades air quality has improved in SoCal and elsewhere. The time span Yglesias is talking about is 1994-2007.
And if you’d checked my link, you’d see that it improved over 1994-2007, too.
June 24th, 2009 at 4:21 pm
So? Yes, roads are actually expensive, and the actual price is rarely represented. Subsidization of roads has led to a death of trains as a means of transportation and polluted our air to no end.
June 24th, 2009 at 11:19 pm
… On a typical day, of course, the United States has zero train-related fatalities.
This is untrue. It does not take into account people hit by trains. According to this there were 716 fatal incidents accounting for 794 fatalities in 2008. Just one of these incidents (with 24 fatalities) involved passengers. Trains don’t look as safe when you look at total deaths per passenger mile not just passenger deaths.
June 26th, 2009 at 9:28 am
[...] One Story You Won’t See in Today’s Paper: 97 Dead in Car Wrecks (Yglesias) [...]
June 26th, 2009 at 1:34 pm
[...] riders this week has renewed calls for rail safety upgrades and reminders that car travel remains far riskier than transit. But the crash is also shedding light on a problem that goes beyond Washington: tax [...]
June 26th, 2009 at 3:05 pm
“This is untrue. It does not take into account people hit by trains.”
Trains are especially dangerous to suicidal nuts?
June 26th, 2009 at 3:47 pm
[...] riders this week has renewed calls for rail safety upgrades and reminders that car travel remains far riskier than transit. But the crash is also shedding light on a problem that goes beyond Washington: tax [...]
June 26th, 2009 at 4:21 pm
[...] riders this week has renewed calls for rail safety upgrades and reminders that car travel remains far riskier than transit. But the crash is also shedding light on a problem that goes beyond Washington: tax [...]