Fatalities per capita vs miles driven

One thing I find amusing about traffic fatality statistics is that it shows how distorted peoples' perceptions of threats are. Using 10-15 as the estimated number of traffic fatalities per 100,000 people, the number of traffic fatalities one would expect among an NYC/LA/Chicago-sized group of people is on the same order as the number of people murdered in those cities every year.

Moreover, unlike murder traffic accidents don't discriminate. If we look at just whites (who in Chicago account for 4% of murders but 31% of people), the number of people dying from murder is about 1/5 as many as one would expect in a similarly-sized population to die from car accidents.

Indeed, the people we care the most about protecting, our precious middle-class kids, are much more likely than average to die from a car accident, while they are far less likely than average to die from inner-city violence.

Statistics like these really call into question those who move to the suburbs because of perceived safety. I grew up in a suburb, and drove everywhere like suburban teenagers are wont to do, and with that in hindsight I'm raising my daughter in a city where she doesn't have to ever get behind the wheel.

That visualization has an Edward Tufte feel about it. I wonder if he influenced the author at all. It's pretty cool.

"Americans drive a staggering number of miles — close to three trillion every year, according to the government. (That is half a light-year, or 120 million trips around the world.)"

It's not often that you see light-years used as a unit associated with driving!