It seems fatalities per 100 million VMT reached a minimum in 2014 (1.08) and has crept back since then [1] (to 1.19 in 2016, 1.13 in 2018), still not much above

It seems fatalities per 100 million VMT reached a minimum in 2014 (1.08) and has crept back since then [1] (to 1.19 in 2016, 1.13 in 2018), still not much above 2012 levels (1.14). I can only speculate that the decline has to do with more safer models on the road and then distracted drivers perhaps bumped things up very slightly again. I think most drivers, at least those not always distracted, notice the prevalence of distracted drivers on the road - cars that drive normally and then, say, gradually slow down to well below the speed limit and the speed of traffic. Which is to say, considering how common they are, it's remarkable these drivers don't cause many more accidents - statistics altogether seem to say they don't (correct me if I'm missing something). A factor to consider is that the "not as dangerous as you think" quality of distracted drivers comes because they do drive slowly, slowing traffic and thereby reducing accident totals (years ago, the introduction of the 55 speed limit reduced traffic fatalities a lot, see same wikipedia article). Basically, the distraction trends seems to have made driving more annoying but not that much more unsafe. Perhaps something still needs to be done but that something might just some warning bells when driver driving outside the norms. [1] https://en.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in_U.S._by_year https://en.wikipedia.org/wiki/Motor_vehicle_fatality_rate_in...