Which is safest? Air or ground?

January 28, 2006 3:16 PM

Is it true that commercial airlines are safer than cars or is this just one of those things that people say?

The statistics that I've seen compare deaths/100,000 miles traveled. According to this statistic airlines are safer; however, this statistic isn't meaningful to me because 100,000 miles to a Boeing 747 translates to a few days of travel (non-stop) while it would take months to travel the same distance in a car.

It is also meaningless to talk about the total number of auto fatalities per year and compare them to airline fatalities. We all drive hundreds of times more often than we fly).

The most meaningful statistic, to me, would be fatalities per HOURS driven compared to fatalities per HOURS flown on an airline. Is there such a statistic? If not, is there some other way to make an apples to apples comparison that would shed light on this question?

posted by crapples to travel & transportation (30 comments total)

If you estimate the average speed of each mode of travel, you can convert from deaths/mile to deaths/hour...
posted by mbrubeck at 3:18 PM on January 28, 2006

The statistics that I've seen compare deaths/100,000 miles traveled. According to this statistic airlines are safer; however, this statistic isn't meaningful to me because 100,000 miles to a Boeing 747 translates to a few days of travel (non-stop) while it would take months to travel the same distance in a car.

True, but this is the stat you need if you need to travel from Boston to Philadelphia and want to know if it's safer to fly or drive. That's the real life question to which the relative safety of each is relevant.

As for deaths per hour, do what mbrubeck said.
posted by duck at 3:22 PM on January 28, 2006

I was fairly sure that deaths/hour are almost equivalent, but I am having trouble finding a citation.
posted by jenovus at 3:28 PM on January 28, 2006

According to this page, every trip in a full-size commercial airliner is about six times less likely to be fatal than a trip in a car.

That seems more useful to me than an hour or a mile-based comparison. I'm not sure what assumptions are made by the NTSB on their automotive trips, but 6x is a sufficient margin, that commercial air seems like the clear winner.
posted by I Love Tacos at 3:30 PM on January 28, 2006

My reason for disliking mile-based comparisons, is that they unfairly favor the airplanes (since the takeoff and landing are disproportionately dangerous).

I dislike the hour-based comparisons because they unfairly favor the automobiles (since the number of hours to make a given trip in a car is much higher than the time required via car).
posted by I Love Tacos at 3:32 PM on January 28, 2006

Sorry for the third post, but the third statistic in favor of airlines is non-fatal wounds. Cars have a huge number of accidents which cause moderate to severe injury, but not death.

Commercial airliners very rarely have accidents in that range. Their accidents skew towards very minor and very severe, with very little in the middle.
posted by I Love Tacos at 3:34 PM on January 28, 2006

My reason for disliking mile-based comparisons, is that they unfairly favor the airplanes (since the takeoff and landing are disproportionately dangerous).

I dislike the hour-based comparisons because they unfairly favor the automobiles (since the number of hours to make a given trip in a car is much higher than the time required via car).

Good points, though I think the per trip favours the airplane. Pretty much all airplane trips are long distances, but most car trips are a couple of miles to the supermarket or office. Surely the longer the car trip the greater the risk, so using average car trip pulls the risk for cars down.

Ideally you'd want given what you said is some sort of measure that had a "fixed costs" of death per trip plus a per mile or per hour co-efficient for each mode.
posted by duck at 3:39 PM on January 28, 2006

Boeing, the commercial airline builder, says that it is 22 times safer to fly than it is to drive on a per-mile basis.

Boeing 747 =825 Km/h = 515 miles/hr
Typical car speed is probably 50 Miles/hr or less

So the plane is covering about 10X as many miles in one hour, so the 22 times safer figure becomes 2.2 times safer on a per hour basis.
And of course the taxi ride to the airport could thow that figure out even further.

I think time is a reasonable way to measure this, people are far more likely to do a journey that takes one hour than 15. e.g. planning a weekend trip the choice would be between flying somewhere distant or driving for a similar time to somewhere much nearer.
posted by Lanark at 3:43 PM on January 28, 2006

I think time is a reasonable way to measure this, people are far more likely to do a journey that takes one hour than 15. e.g. planning a weekend trip the choice would be between flying somewhere distant or driving for a similar time to somewhere much nearer.

I suppose I should also apologize for making a third post but..

That's interesting because it brings up how the different way people travel would affect their evaluations. "I want to go away this weekend should I go near or far?" would never ever occur to me. Normally I have somewhere to go (that's given) and it's a matter of how to get there. And I rarely go anywhere for the weekend. I'm much more likely to go somewhere for 2 weeks to a month (or longer). If I were just picking a place to go and it wasn't already given, I would for sure choose somewhere far. So for me the "per trip" (or the modified per trip described above) would be much better then per hour.
posted by duck at 3:52 PM on January 28, 2006

Surely the longer the car trip the greater the risk, so using average car trip pulls the risk for cars down.
Not that much. Think: short trips in a rush-hour/typical traffic with lots of intersections and stop lights vs. long trips on interstates where you're just driving straight.
posted by martinX's bellbottoms at 4:17 PM on January 28, 2006

The rate per mile is the obviously fair comparison. What decision would you need to make that would require a different measurement?

"I've got a few hours to kill. I think I'll just go for a little drive around town.... no wait, I'll fly around for a few hours instead! But wait, is that safe?"
posted by crabintheocean at 4:18 PM on January 28, 2006

Not that much. Think: short trips in a rush-hour/typical traffic with lots of intersections and stop lights vs. long trips on interstates where you're just driving straight

My young drivers' instructors said city driving is far more likely to result in accidents, but accidents on the highway are much more likely to result in fatalalities. I still think the trip to the office and supermarket pulls the average number of deaths per trip down for cars.
posted by duck at 4:19 PM on January 28, 2006

National Transportation Statistics
posted by milkrate at 4:25 PM on January 28, 2006

I believe the Federal Aviation Authority has refused to required infants to have their own airline seat, although that would be safer (the infant could be buckled in, along with a baby chair). The rationale is that (a) airlines are extremely safe right now; there would be little savings of life by requiring this; (b) if a separate seat HAD to be purchased, parents would pay more to travel by air; (c) some parents would be disuaded by the higher airline cost, and would travel by car instead; (d) cars are much more dangerous (more than 30,000 deaths per year in the U.S.; far more injuries); (e) the requirement would therefore lead to MORE deaths of infants (traveling in cars) then it could possibly save in (safer) airline conditions.

In short, if you have to decide between driving to a location, and flying to a location, and safety is a major concern, then FLY. There hasn't been a major airline crash (as in, over 100 people in the plane, and most people die) in North America for four straight years, per this Salon (subscription required, sorry) story.

One reason you may think that airline travel is dangerous is that every time there is a minor mishap, it's on network news. On the other hand, when was the last time you saw a front page story in the New York Times or Washington Post, or national network news report about a car crash that killed (say) four people (people who weren't famous)? Such car crashes and deaths occur every day in the U.S. They aren't newsworthy precisely because they are so commonplace.

A lot of people seem to think they're safer in their own cars because they're driving - they're in control. Yes, but they are not in control of the drunk driver coming at them from the other direction, or the inexperienced teenage driver at the upcoming intersection, or the nail in the road that is about to puncture their tire. In fact, they'd be a lot safer in an airplane piloted by folks with years of experience, in a controlled environment (airport, air space), where everyone (pilots, mechanics, air traffic control, airplane manufacturers, airlines) are absolutely concerned with avoiding serious accidents.
posted by WestCoaster at 4:27 PM on January 28, 2006

This is not a wisecrack ... call it an anecdotal example. This is what I tell my mother when she's terrified of flying.

Think about how many people you personally know that have been in car accidents. Dozens, right? Probably every single person you know has been in at least a fender-bender.

Now, how many people do you know that have been in plane crashes?
posted by frogan at 4:31 PM on January 28, 2006

However, Frogan, this is exactly why a per hours or per miles comparison is needed: We drive so much more often than we fly that even if driving was slightly safer than flying you'd still expect to know more people who have died or been injured in cars.

From the links and statistics that have been provided, I think it's safe to say that flying is safer than driving on both an hour by hour basis and a mile by mile basis.
posted by crapples at 4:37 PM on January 28, 2006

The statistics that I've seen compare deaths/100,000 miles traveled. According to this statistic airlines are safer; however, this statistic isn't meaningful to me because 100,000 miles to a Boeing 747 translates to a few days of travel (non-stop) while it would take months to travel the same distance in a car.

those statistics also don't take into account the precipitous increase in the number of fatalities as cars reach cruising speed and/or altitude: i believe that driving or riding in a car travelling at 500mph at 30k feet nearly always results in death for all occupants, but i don't have any references.
posted by herrdoktor at 4:37 PM on January 28, 2006

I hate flying, I'm paranoid, I get nervous, I have death fantsies, I pray on takeoff. Then I relax with the tought the I don't know anybody who has been in a fatal airliner accident, I only knew one person who parents had died in a small aircraft crash (when I was in High School) and I know lots of people who fly a lot more than me (I fly about fifteen times a year) who have never had accidents.

On the other hand I know lots of people who have had car accidents, several who dies this way, and had about four good accidents before I up driving cars when I was 20. I shit myself in cabs to the airport. Many, many years later, however, I am still alive enough to pee myself on airplanes.
posted by zaelic at 4:51 PM on January 28, 2006

frogan; yes, you probably do know more people who were in car accidents. There aren't that many plane "accidents" though. When a plane crashes you and everyone on that plane are going to die, probably.
posted by odinsdream at 5:35 PM on January 28, 2006

Note however, that I'd fully support much, much tougher licensing requirements for operators of cars. Probably not as much as that for pilots, obviously, but definitely more than being able to spell your name correctly and fill in some bubbles.
posted by odinsdream at 5:37 PM on January 28, 2006

This came up in the blue, and from the figures cited, cars are safer than planes per hour.

However I didn't double check the figures.
posted by -harlequin- at 5:40 PM on January 28, 2006

Direct comparison is difficult, because those activities are actually quite different. Driving is basically a day-to day continuous activity and flying is more of a matter of taking distinct trips. Let us then find out how many miles of driving (or being a passenger) is as risky as a single trip in an airplane. Let's try another kind of comparison: Airplanes average 0.019 Fatalities per million trips where-as cars average 22.5 fatalities per 100.000 license holders per year. If my math is correct, this means that One week of normal driving is roughly as dangerous as 200 plane trips.
posted by insomnus at 6:07 PM on January 28, 2006

Note however, that I'd fully support much, much tougher licensing requirements for operators of cars. Probably not as much as that for pilots, obviously, but definitely more than being able to spell your name correctly and fill in some bubbles.

A counterintuitive option is to take the Motorcycle Safety Foundation classes. Not only do you get to ride motorcycles in a controlled environment, but they teach you a LOT of useful ways to avoid getting killed while riding one. Almost everything they suggest translates well to cars.
posted by b1tr0t at 6:22 PM on January 28, 2006

My reason for disliking mile-based comparisons, is that they unfairly favor the airplanes (since the takeoff and landing are disproportionately dangerous).

To be precise: by far the most accident-prone part of a "flight" is the taxiïng on the ground.
posted by Aknaton at 9:32 PM on January 28, 2006

Steven Levitt's Freakonomics answers this question. The "meaningful statistic" that you chose, crapples, is the same one that Levitt uses -- the per-hour death rate of driving versus flying. An excerpt:

If you are taking a trip and have the choice of driving or flying, you might wish to consider the per-hour death rate of driving versus flying. It is true that many more people die in the United States each year in motor vehicle accidents (roughly forty thousand) than in airplane crashes (fewer than one thousand). But it's also true that most people spend a lot more time in cars than in airplanes... The per-hour death rate of driving versus flying, however, is about equal. The two contraptions are equally likely (or, in truth, unlikely) to lead to death. (151)

posted by medpt at 9:34 PM on January 28, 2006

Aknaton: Taxi accidents occur at low speeds (15-20mph) and incur damage to the plane and to the aircrew's careers.

Takeoff and Landing are "critical phases of flight" where a mishap has great potential to be disasterous.

In flight problems are generally not too serious because even if all the engine die you can still glide down (However, airborne emergencies can certainly be fatal).

Remember, aircraft undergo constant maintenance to prevent mechanical failure. They have highly trained crews that go to simulators and practice different emergencies so they can respond instantly and accurately to a crisis. In flight and on the ground you have external agencies watching what you're doing to try and prevent mishaps. Do any of these positive factors apply to driving?
posted by tcobretti at 9:54 PM on January 28, 2006

Is it true that you can glide down if all the engines die? Those airplanes of today look more like bricks with a jet motor to me. I mean: they're so overpowered that even a small flap can get them airborne. But take away the power and they go right down [/uninformed anxiety speaking].
posted by NekulturnY at 4:16 AM on January 29, 2006

Yes, it's true that you can land a 767 as a glider if all the engines fail (true story)!
posted by mbrubeck at 8:02 AM on January 29, 2006

It seems intuitive to me that one's chances of dying (or, more importantly, being seriously injured) in a road trip from New York to LA are far greater than than spending 6 hours on a flight.

While it's true that we individually spend more time in cars than we do on planes (unless you're like me and spend most of your time on busses and trains), bear in mind that an individual flight might have 200 passengers. This would equate to 1200 person hours per NY to LA flight. This equates to 50 days of travel for a single person in a car, 10 days for five occupants, etc.

From this perspective, I'd think it far safer to fly.

I'm not particularly comfortable on planes, but I've somehow managed to get used to it. There was a time when I'd get sick before each flight and panic at the slightest turbulence. Now, I'm the guy who casually reads the paper while the plane jumps. It's almost always smoother than my morning commute via Chicago Transit Authority.
posted by aladfar at 9:31 AM on January 29, 2006

It also depends on whether you're driving a car or motorcycle, or driving in Mississippi or Massachusetts. The really bad news: as you get older, you're even more likely to die as a pedestrian.
posted by weapons-grade pandemonium at 12:04 PM on January 29, 2006