From a Wall Street Journal article today on federal fuel economy regulations, and the perverse incentives they create for auto makers:

While light trucks represent 36% of all registered vehicles, they are involved in about half of all two-vehicle crashes with passenger cars, highway safety regulators say. In such crashes, over 80% of those killed were in passenger cars.

So if you bought that SUV for “safety,” just keep in mind that you bought that safety at the expense of the safety of others.

And if you bought it just so you would look cool … well, you don’t.

UPDATE: James at OTB correctly points out that the statistics I cite don't make it clear to what degree the 80/20 death ratio in SUV-car crashes is a matter of a higher fatality rate in cars vs. a higher survival rate in SUVs, comparing with car-car crashes.

The source for the WSJ factoid wasn't hard to track down. It's from the Feb. 23 testimony of the Administrator of the NHTSA before a Senate commitee. Here's a better statistic that the WSJ should have used, one which does support my thesis:

When controlling for impact location, and comparing light trucks to passenger cars of comparable weight, we found that light trucks were more than twice as likely as a car to cause a fatality when striking a car.