I’m going dataless for the moment, but I suspect my guesses here are close. If you were to, right now, take 1 additional trip in your car and compare the risk of injury or death to taking 1 additional trip on a bicycle and 1 additional trip on a subway or 1 additional trip on a bus or 1 additional trip on an airplane, your chances of dying and injury would be largest by driving in a car. Holding constant the obvious reasons why people may prefer one or the other, if you actually cared about the safety of people in the transportation sector, would it not make sense to (in addition to promoting car safety), purposefully make airplanes crash more often? Similarly, wouldn’t we want to see buses and trains and bikes and boats become more dangerous? As long as, at the margin, those modes of transportation will still be safer than car driving, then by making those other modes more dangerous you would be making them cheaper, and therefore transitioning some people from their cars and onto the now slightly more dangerous than before (but still safer than cars) other modes of transportation?
The logic here is undeniable. I am sure it is not something you want to suggest in conversations at your next dinner party should you want to be invited to another one.