Recently, David Stokes of the Show-Me Institute made the surprising statement that, per passenger mile, cars are safer than trains and using his claim as an argument against building high-speed rail in the United States.

Later in the interview, Harnish gives a great little aside downplaying safety of cars and claiming, by insinuation, that trains are safer.

. . . if you believe that our strength and unique identity is tied to the ability to risk your life everyday in a car . . .

OK, so we risk our lives everyday in a car. Would we not risk them in a train? Now, I am not saying passenger trains are unsafe — they are indeed safe. But if you compare them to cars, there are more fatalities on passenger rail than in motor vehicles per passenger mile. According to the latest data, passenger cars have 0.9 fatalities and 83 injuries per 100 million passenger miles. Passenger rail has 2.9 fatalities and 1,226 injuries per 100 million passenger miles. So they are both safe, but let’s not pretend passenger rail is safer.

It’s great that David is using and citing real data when making his claims on the safety of motor vehicles. The problem is that he’s wrong. Very wrong.

David’s mistake is from misinterpreting the units used in the data source, confusing passenger train miles and vehicle miles with passenger miles.

What’s the difference? Imagine you have a 5 car Amtrak train with 25 passengers per car traveling 240 miles between St. Louis and Kansas City. That trip would entail 240 passenger train miles, 1,200 vehicle miles, and 30,000 passenger miles.

According to the data used by the Show-Me Institute, there were 2.9 fatalities per 100 million passenger train miles and 1.1 fatalities per 100 million vehicle (car) miles. However, using a different data set that includes passenger miles by mode, from 2005 to 2009, there were 1.8 fatalities per billion passenger miles on intercity trains while there were 8.3 fatalities per billion passenger miles on highways. Injury data is similar.

Whoa! Big difference! Big mistake!

And the opinion at the beginning of the Show-Me Institute’s article that statements made by Rick Harnish [of the Midwest High-Speed Rail Association] are misleading, I believe, is misleading itself.

Posted by Herbie Markwort

I like to write about transportation.


  1. Nice work. I've been waiting for the Show-Me Institute to fix that on their blog for a few days now and I don't think they're going to.

  2. Yes, that's exactly why I posted this. I was tired of waiting for them to correct their glaring mistake.And I was being generous by picking the 2005-2009 timeframe as there were 3 major train accidents in that range. A website I found managed by AASHTO claims that there have only been 0.4 fatalities per billion passenger miles on Amtrak over the last 30 years compared to 12.9 fatalities for autos.

Comments are closed.