> Deaths per 100,000 people is probably the wrong figure to look at to determine the safety (or lack thereof) of driving.

Kind of. It's useful to see the per distance and per population figures. You are likely to drive more when you are in the USA compared to Europe (due to the culture and bigger distances to travel) and this increases your risk.

If I was analyzing my total risk of dying in a road accident, both figures would play a part.

I agree. However, I find that even more remarkable. I looked at the wiki statistic yesterday and the comparison between Germany (where I grew up) and the US made me think. Roads in Germany do feel safer and better organized. But because of the famously high speed limits on the Autobahn I didn't expect to see number like this. I used ot go 220km/h in a Golf 2 from 1994. That was even scarry back then at times and seems downright insane now that I am used to US traffic.

As someone else mentions, I think that drivers pick lanes at will leads to constant lane changes and also to traffic jams and reduces throughput. That to me is a huge thing that needs fixing. It also makes driving much more pleasant. You never have a "wall" of three slow cars in front of you and now have to figure out which one of them will eventually make way, so that you can pass and now everyone has to change lanes. It's insane! In addition to the broken "protocol" that's being used on roads cars are also much less save. Lots of pickups have these pipes on front of their car's grills (sorry no idea how those things are properly called). These are actually illegal in Germany because they increase injury likelihood for pedestrians. Trucks in Europe have to have barriers on the side, so that in a collision they push a car aside and have a big impact area, rather than just hitting with a car with the edge of the trailer just at the same height as where people's heads are. Really cheap to do and would save lots of lifes. There are also many more very large cars in the US. If those crash into a smaller car that can have much more severe consequences. I think the lack of monster pick up trucks in Europe helps. I know many people here get these trucks to defend themselves against other people who are in these huge vehicles.

That figure isn't particularly illustrative either, conisdering that, in much of the US, living a normal life requires driving significant distances on a daily basis. In most other countries, driving is an optional convenience.

The deaths per vehicle-kilometer figure doesn't illustrate that it's very difficult to opt out of the automotive dangers in the US. And thus we also have a hard time forcing dangerous drivers to stop driving, in order to keep the rest of us safe from them.

I disagree. Comparing two countries with the same deaths per billion miles, the one where geography and urban planning causes people to drive more miles is more dangerous.

What about a metric that accounts for speed? Should it account for that?

Intuitively, if two cars each average one death per million miles, but one of them had an average speed of 5mph, and the other 60mph, then the faster one should be regarded as safer because it maintained that death rate under less safe conditions.

That would help, I think. I suspect a difference in average speed is partially behind the differences between Montana and New Jersey, possibly more important than the usual "Montana drivers are freedom-loving (or insane) and don't like to be told to wear seatbelts."

But in Japan, people generally take public transit for ordinary travel and only use their car for extraordinary travel. I suspect the death rate of "deaths per billion vehicle-kilometers of extraordinary travel" in the states would be much higher than 7.6.

Not quite so sure...

1. Deaths per 100,000 seem to be pretty good a measure for pedestrian deaths

2. Your measure does not consider the cost of driving more

But, one point I think you have: If indeed the number of deaths per km is similar, then the solution to car deaths is to drive less, not to drive more safely

It seems as if the author had a point they wanted to prove and then tried to find some facts that lined up with their point.