> They're not all contributing to that specific accident, but they're contributing to the rate of accidents. > For a more extreme example, imagine a million dr

> They're not all contributing to that specific accident, but they're contributing to the rate of accidents. > For a more extreme example, imagine a million drunk drivers getting on the road at the same time to drive a significant distance. Nobody has crashed yet, but many of them will. I would say they are clearly "contributing to causing fatal accidents". Would you not? It's like you are saying Shark Attacks are an externality of swimming in the ocean. That is not an externality, that is a risk of swimming in the ocean. Someone swimming in the ocean and is fine is completely independent of someone else who swims in the ocean and is attacked by a shark.