Suddenly, flying seems scary again.
After months of harrowing near-misses on U.S. airport runways, an actual runway collision took place in Tokyo this month. While all passengers on the large plane survived, it was frightening to see color photos of an Airbus 350 engulfed in flames.
And while the rapid decompression on the Alaska Airlines MAX jet caused no serious injuries, we were informed that the event could have been catastrophic had it occurred at a higher altitude.
Media reports about the event routinely mentioned prior MAX crashes had caused 346 deaths. Moreover, terrorism experts have expressed the fear that the Israel-Hamas war could stimulate new attacks around the world — attacks in which aviation is often a prime target.
At times like this, it can seem exasperating when statistics are cited to suggest that passenger flights are extremely safe. But such statistics are valuable because they consider air transportation as a whole and not just a small set of highly publicized events. Thus, I ask the reader’s indulgence as I present some data below about the mortality risk of passenger air travel.
The chances of getting killed during an air journey has always been low, and has relentlessly been improving. For the period of 1968 to 2022, the worldwide risk for a randomly chosen passenger on a scheduled flight has been as follows:
1968-77: 1 in 350,000
1978-87: 1 in 750,000
1988-97: 1 in 1.3 million
1998-2007: 1 in 2.7 million
2008-17: 1 in 7.9 million
2018-22: 1 in 13.4 million
This pattern is approximated by a simple rule of thumb: each decade, death risk per flight fell by a factor of two. That means between 1968-77 and 2018-22, the chance of dying in flight has dropped by a full factor of 38!
One would also think that, as the risk gets lower, the harder it would be to keep reducing it at the same rate. But, astoundingly, the world’s airlines/governments/aircraft manufacturers have managed to do so.
The story is even more dramatic for scheduled U.S. airlines, which have always been among the world’s very safest. Based on their death risk per flight for 1958-62, one would have expected a total of 12,470 passenger deaths on those airlines in 2018-22. At the risk level for 1978-82, there would have been 1,850 deaths in the most recent period; using the 1998-2002 numbers, there would have been 700. The actual number of passenger deaths over 2018-22? Eleven.
Those 11 deaths arose among the 3.7 billion passengers carried over 2018-22 on scheduled U.S. airlines. That works out to a death risk per flight of 1 in 337 million. At that rate, a passenger could on average take one flight every day for 922,000 years before succumbing to a fatal crash.
When the risk of flying is so minuscule, being afraid to board an airplane is hardly more justified than avoiding the supermarket for fear that the ceiling will collapse.
And recent developments might actually make us safer still. Compared to a few weeks ago, we are far less likely to grab a carry-on bag during an emergency evacuation. We understand better why it is essential to keep our seat belts fastened when seated. And, after years when we could practically forget about terrorism, we realize once again that if we see something, we should say something.
The safety of flying in countries like the U.S. is the eighth wonder of the world. Far from being nervous as we approach the airport, we should be awestruck that flying is so free of risk — and deeply grateful to those who have made it so.
Arnold Barnett is George Eastman Professor of Management Science and a professor of statistics at MIT Sloan School of Management.