About the data
The data used here is from both the NHTSA's Fatal Accident Reporting System, and the FHWA.
The ranking itself is actually very basic: it's the total population of each state divided by the number of people killed in automotive accidents for the given year (2013).
The rest of the data involved, however, is where things get interesting:
Being Involved in a fatal crash: The NHTSA keeps such thorough accounts of every fatality accident that we know exactly how many people were in each vehicle for every single crash. Since surviving such an incident is inarguably a horrific and traumatizing experience, dividing the total population by the number of crashes is a sobering way to realize just how fragile life can be.
Being involved, as a driver: Simply put, the more you drive, the higher your chances of being involved in an incident. We took the population of licensed drivers and divided by the total number of accidents.
Dying in a crash, as a driver: Similar to being involved in a crash, because the NHTSA actually notes who was injured or killed while sitting where in a vehicle, we were able to calculate the odds of dying as a driver by dividing the licensed population by the number of people who died while driving.
Dying while you're not even in a moving vehicle: This includes pedestrians, bicyclists, people inside buildings that get driven into, and parked cars. Basically, anyone who's not out in traffic in a vehicle. As you might expect, it's the total population divided by the number of people who were struck and killed.
There are some problems with the data: While the statistics clearly show a trend toward safer urban driving, they are skewed in areas with well-developed mass transit. The percentage of people who have a license and don't drive is not accounted for in the data, and would potentially skew a few areas where car use is less prevalent.
It's more than a little somber to realize that every one of these numbers was caused by the death of someone's spouse, parent, or child. Still, the data is fairly conclusive that rural driving is far more likely to kill you than urban driving -- and the FHWA has the specific breakdowns of this on its site. This data didn't look into exactly why that is, but issues like proximity to first-rate medical care and vehicle speed -- since you're not crawling along in traffic -- are likely greater factors in rural areas.
Interestingly, the number of vehicles per crash doesn't appear to bear significant correlation to how many people die in each crash. Presumably, that number is more dependent on other factors (weather, intoxication, time of day), than it is on the violence of the collision itself.
Want more of the world's best Cars delivered straight to your inbox? Click here to sign up for our daily email.
Aaron Miller is the Cars editor for Thrillist, and can be found on Twitter. His closest call was during a road trip 15 years ago, on a rural stretch of I-35 in Texas.