cars had very few safety features available today. Nobody is going to argue 1982 cars are safer. Where is the control group data to account for this?
It doesn't need control group data, it's a percentage of all auto fatalities. Cars ARE safer than in 1982 as is evidence by the almost 7000 fewer fatalities in 2008 compared to 1982. What you're missing is that from 1982 to 2008
alcohol-related fatalities were cut almost in half. The rate of alcohol-related fatalities dropped faster than the rate of all auto fatalities. Vehicle safety is already accounted for.
Voluntary breathalyzers wouldn't work. Most DUI/DWI offenders, especially repeat offenders, don't believe they have a problem. They won't voluntarily purchase one, so you get adverse selection.
The only way to stay safe on the roads is to stay off them. You can legislate and regulate to your heart's content and you will never make them 100% safe. There is an assumption of risk when you get in an automobile and it's an assumption you
voluntarily make. DUI/DWI is already illegal and enforced but you assume the risk that someone might not be caught just like you assume the risk that someone might apply makeup, eat a hamburger, talk on the phone, speed, not stop at a light or sign, cut you off, brake suddenly, blow a tire, not see you in a blind spot, or any other of the myriad risks that face you when you drive or ride in a vehicle.