It depends what you're trying to measure, no? I agree that for the point being made it's not ideal, but I think it's separately noteworthy that in the US, poor planning and more car-dependent lifestyles lead to a higher rate of traffic fatalities because of how much more Americans end up having to drive.