Reader Kevin Wenzel raises the point in email that, contrary to what might be inferred from one of my posts, the US death rate per vehicle mile travelled is only marginally worse than that in Australia. I actually addressed this in my Fin article, but since this is virtually inaccessible, I’ll state my points here.

I have three problems with deaths/VMT as the measure of road safety.,
(a) It doesn’t take account of vehicle occupancy. So a car with passengers counts the same as a car with driver only, although more people are travelling and therefore at risk of death or injury
(b) It’s not relevant in relation to risks to non-motorists: from an economic or libertarian viewpoint these are of particular concern since they’re involuntary externalities
(c) It takes car-dependence as given. This is a complex issue, but it’s striking that distance travelled/vehicle has remained almost constant in Australia, a country with similar geography, population growth and income growth, while in the US it has grown strongly.