Sean Carroll (who's moved, by the way, over to Cosmic Variance) takes a closer look at the claim that airplanes are safer than cars and concludes that a lot of information can be hidden by statistics (cf lies, damn lies). His discussion reminds me of the claim my wife frequently makes that you're more likely to get in an accident close to home (the context for which is my laziness in putting on my seatbelt for very short trips). I always respond that if this claim is even true, it's probably just the trivial result of the fact that most of your trips either begin or end at home. Of course, my response isn't much of an argument for not wearing a seatbelt close to home, since that would mean a very large number of seatbeltless trips. And the whole discussion kind of skirts the issue of seat belts and short trips.
September 16, 2005
I think I've always heard the safety quota expressed as fatalities per passenger mile. Airplanes cover a lot of space in a very short time, so it wouldn't surprise me if the fatalities per man-hour of travel number for each mode was very different from the per mile.
For that matter, I wonder if space travel is safer per passenger mile than anything else. Those are some pretty amazing distances involved.
Yeah, I'm not sure there's a big enough sample to talk about space travel. All of these kinds of calculations require a huge sample, because we're talking about an extremely low probablility event that's extremely bad. It's the same reason that policy makers have such a hard time dealing with cataclysmic hurricanes.
Post a comment