That commonly repeated datum is one of those that falls under the lies, dirty lies, and then statistics - type of lies. Because most people would think this means that if you instead had a private jet and just started flying literally everywhere (instead of driving) then it'd be much more safe, right? Nope - it'd be dramatically more dangerous in fact.

This owes to the fact that the stat compares deaths per mile, but the real risk in a plane is not measured in miles, but in takeoff/landing/pressurization cycles. Whether the plane then flies for 10 miles or 5000, the risk is pretty much the same. And so if you mask out the danger by looking at deaths/mile then yeah - they're safer. But if you look at it in terms of 'what are my odds of dying on an average trip' - then they're much less safe than driving.

You can really intuit this by imagining some ultra-high risk interstellar starship. If it traveled quadrillions upon quadrillions of miles, but 99% of them randomly blew up and killed everybody on board, then it'd still be far safer than driving in terms of deaths/mile. The deaths/mile datum is not really very useful for modes of transport where the risk models are very different, and that's very much the case for driving vs flying.