With the loss of at least eight lives, this week's train crash in Philadelphia is of course a national tragedy. It's crucial to understand why the accident occurred and to take steps to make such tragedies less likely in the future. But there is also a serious risk of overreaction. Many Americans might feel inclined to avoid taking trains, even though their safety record is extraordinarily impressive.
There are two reasons for such overreactions, both of which have been carefully explored by behavioral scientists. The first is called"probability neglect": When a horrific outcome makes people's emotions run high, they focus their attention on that outcome, without thinking much about the (low) probability it will actually occur.
People tend to neglect probability when they are focused on highly desirable outcomes, such as winning the lottery or kissing someone they really like. They also neglect probability when they are focused on catastrophic outcomes, such as dying in an earthquake or a train accident. The problem is that if people don't keep probabilities in mind, they might forgo beneficial activities in order to avoid risks that are extremely small.
The second source of public overreactions is called the "availability heuristic." First identifiedby Nobel Prize winner Daniel Kahneman and his coauthor Amos Tversky, the availability heuristic is a mental shortcut, or rule of thumb, through which people assess the likelihood that a risk will come to fruition. Instead of looking at statistics, we ask: Can we think of a situation in which a risk actually materialized? That isn't the most foolish question to ask, but it can lead to big mistakes, including unjustified fear as a result of recent or memorable events. Even if there was a theft in your neighborhood last month, the risk of theft, for you, might nevertheless be pretty close to zero.
In fact, that's the situation we now find ourselves in with trains. One careful study of transportation risks from 2000 to 2009 showed that, per one billion passenger miles, riding a motorcycle is by far the most dangerous mode of transportation, accounting for 212.57 deaths. Driving or riding in a car or a light truck is far safer, with 7.28 deaths. Being a passenger on a local ferryboat is safer still, accounting for 3.17 deaths. With trains, subways, buses, and commercial airlines, the risks get much lower, ranging from 0.43 deaths on trains to .24 on subways, 0.11 on buses and 0.07 on planes.
The difference between trains and planes is significant, but 0.43 deaths per one billion passenger miles isn't much by most measures. (And in recent years, Amtrak's accident rate has been getting significantly lower.) For perspective, the National Weather Service estimates the annual risk of being struck by lightning as 1 in 960,000 -- in other words, if you ride one billion miles on passenger trains in a given year, you'd still be significantly more likely to die from a lightning strike than a derailment .
Any preventable death is a tragedy, of course. And if (as now seems possible) excessive speeds were responsible for the Philadelphia crash, then we need to take steps to prevent trains from hitting such speeds in the future. But whatever behavioral science says, let's hope travelers don't overreact. And on that score, we can all do our part. I'm planning to be in Philadelphia next week, and if the trains are running, I'll be on one of them.
Cass R. Sunstein, a Bloomberg View columnist, is director of the Harvard Law School's program on behavioral economics and public policy.