Arnold Barnett, a statistician at the Massachusetts Institute of Technology who has studied airline safety, tells NPR that from 2018 to 2022, the chances of a passenger being killed on a flight anywhere in the world was 1 in 13.4 million. Between 1968 to 1977, the chance was 1 in 350,000.
"Worldwide flying is extremely safe, but in the United States, it's extraordinarily so," Barnett said.
In the U.S., there has not been a fatal plane crash involving a major American airline since February 2009, though there have been a handful of fatalities since then.
Brickhouse, who has studied aviation safety for over 25 years, often tells people that the biggest risk of any air journey tends to be driving to the airport.
More than 40,000 people are killed on U.S. roads each year.
"Aviation remains the safest mode of transportation," he says.
Forget how many people are killed, because when you get to that point, then things have already gotten really bad. Look at the leading indicators first: how many near-misses and other incidents (like mechanical failures) have happened in the US over the last 25 years, and is the trend up or down?
The incident with the 737 door falling off is a good example here: this would have been a fatal incident if this had been a full flight. Thankfully, the seat next to the door was empty that day, so no one got seriously hurt, but it could have been much worse.
It's hard to tell for sure without a reliable source of unbiased data rather than various news stories, but it sure seems that the frequency of incidents (in the US) is rising lately, not falling, and that's not good, it's like the canary in the coal mine. Things need to be fixed before planes start falling out of the sky with spectacularly fatal results because too many deep systemic problems have come together to destroy the safety record that existed before. Instead, too many people want to rest on past successes, saying "look! It's so safe compared to driving!" and do nothing.
Crackheads jumping in front of buses has nothing to do with whether motor vehicle travel is safe. An intellectually honest comparison would be to compare the incidence of fatalities while being driven around in a recent model sedan by a professional driver
This isn’t the conundrum you think it is. My estimated statistical value of life (revealed preference) is ~$30M, so I would get on a flight with a 10^-9 risk without thinking twice, but would value the 10^-6 risk at about $30. That is, I would choose to fly in a plane that had an additional 10^-6 risk of death if it was $50 cheaper, but not if it was only $5 cheaper.
Until it's not, right?
Historically, it has been.
If there was a 1/1,000,000,000 chance you were going to die on a plane ride, would you voluntarily choose to take it? Ok, what about 1/1,000,000?