3 April 2024
“Ah, it’s a Boeing Max,” I exclaimed to my travel companions after we boarded our plane a few weeks ago. I looked to see if we were seated next to a hidden door plug panel like the one that blew out on Alaska Airlines Flight 1282 in January. We weren’t, but joining a trend on social media, we cracked a few jokes at Boeing’s expense: “Maybe they can charge extra, saying it’s potentially an even bigger window seat.”
The Federal Aviation Administration recently informed the passengers on that ill-fated Alaska Airlines flight that they may have been crime victims. The agency hasn’t explained why, but Boeing has told the Senate that it cannot find documentation of exactly how the door plug was removed and reinstalled, even though the company acknowledged it is supposed to have kept such records. Facing all this, the company announced last week that it was replacing its CEO. But the bad news wasn’t over: On Thursday, a New York Times investigation reported a disturbing pattern of sloppy safety procedures and dangerous cost-cutting. One expert who had spent more than a decade at Boeing told the Times, “The theme is shortcuts everywhere — not doing the job right.”
Is it any wonder that some travelers are trying to avoid Boeing planes? Kayak, a travel booking site, noticed an uptick in the number of people trying to weed them out; it recently made that search filter more prominent and even added an option to specifically avoid certain models.
Boeing’s problems, great as they are, are just one reason that consumers might be wary of taking flight. United Airlines now also faces scrutiny for a series of safety incidents, although many experts say the issues there do not appear to be systemic. The biggest danger of all may be understaffed air traffic controllers and overstuffed runways, which lead to far too many near misses.
Personally, I am not worried about flying, and other than cracking some ill-advised jokes, I have not changed my behavior. That’s why I hadn’t bothered to check whether I’d be flying on a Boeing Max, or any type of Boeing plane, until after I boarded.
Checks and redundancies
The trajectory of Boeing as a corporation, however, is another matter. It’s going to take a lot more than a shuffle at the top to fix that company’s problems. But the fact that Boeing managed to cut as many corners as it did is testament to the layers and layers of checks, redundancies and training that have been built into the aviation industry. Aviation safety is so robust because we made it so.
Two seemingly contradictory things are both true: U.S. commercial passenger airlines have gone an astonishing 15 years without a single death from a crash. And there is a huge safety crisis in commercial aviation that we urgently need to fix.
Commercial aviation is a complex system involving many dynamics: technology, engineering, corporate culture, regulation, weather, human factors, politics and more.
It’s extremely hard to predict what will emerge from so many different things interacting all at once — an example of the so-called butterfly effect, in which a tiny insect flapping its wings leads to major weather on the other side of the world. And although testing every part of the system on its own is necessary, it’s insufficient, since it’s the interaction of many moving parts that creates those hard-to-foresee problems. Solving equations won’t be enough to manage it all because such systems defy easy calculations.
Layers of precaution
We do, however, have methods to manage complex and safety-critical systems, and if done right, they can work very well.
Perhaps the most important measure is redundancy, the layering of precautions. Since even a minor failure could set off a catastrophic chain of events, it’s important to shore up everything. That’s why many plane parts have duplicates or backups, and much of planes’ production and maintenance is subject to inspections by multiple people.
Redundancy, however, while great for safety, is expensive.
The first Boeing 737 Max crash occurred in Indonesia in 2018. Everyone on board was killed. The next was in 2019, in Ethiopia. There were no survivors of that flight, either. After that, the planes — which had been flying globally for more than a year — were grounded by the FAA. (About 387 of them had been delivered at that time, and 400 or so more were in production.)
The public later learned Boeing had added a new software system to the planes to help keep them stable. Because the system made the planes behave more like older Boeing models that pilots were already familiar with, the company got permission from the FAA to avoid retraining pilots on the new planes (a cost savings for the airlines that bought them) or even telling pilots about it.
Those two flights proved the danger of that approach. The new system relied on a single sensor, even though the planes were equipped with two. When that sensor failed, pilots lacked the information to diagnose the problem and avoid disaster. Boeing’s actions were a violation of those core tenets of aviation of building in redundancy and understanding how complicated interactions can create problems that no one predicted.
Check everything
Given the impossibility of testing for every outcome, keeping complex systems safe also depends on another crucial signal: near misses. If something goes wrong but disaster is averted, the correct response should not be a “whew” and back to normal. It should be caution and investigation.
The Times investigation shows how alarmingly different Boeing’s approach was.
The Boeing plane that crashed in Indonesia had, the day before, experienced the exact same problem with the new stabilization system. But on that flight happened to be a third pilot, riding off-duty in the back of the cockpit. When things went haywire, he was able to suggest the correct sequence of actions and saved the day. Had Boeing updated pilots about the system, would the passengers on the airplane’s next flight have landed safely? We’ll never know.
That third pilot — in that case, present purely by luck — was an example of how redundancy can save lives. So is a co-pilot. Planes fly on autopilot all the time, and can even land on their own. Still, regulations require a second person in the cockpit for many types of passenger flights, not just to handle things in the extremely rare event that the primary pilot gets sick or dies midflight, but to help manage emergencies and equipment failures. It’s the same reason that planes have more engines, more tires and more ways to extend the landing gear than they need for any individual flight, just in case one of those things fails, as has happened many times.
An extra layer of safety helped avert the Alaska Airlines blowout from turning into a catastrophe: Because the incident occurred so soon after takeoff, all the passengers were still required to wear their seat belt.
Pilots even do a “walk-around” of their plane just before takeoff to conduct a final visual inspection. Commercial aviation works because of the principle of trust nothing and check everything.
It’s hard to escape the conclusion that those at the company who took all those shortcuts figured the system, with all its redundancies, would save them. But that’s a gamble. Eventually, two or three or four rare mishaps will align.
A Boeing representative told me that the company was taking responsibility and working to improve quality. But we need to see action, not promises.
Still, safe. Let’s keep it that way
So, why should anyone still fly on Boeing’s planes? Or fly at all? Because the statistics still show that commercial aviation is miraculously safe, far more so than all the alternative ways of traveling.
Although I don’t check for who manufactured the planes I fly, I do keep my seat belt on even when the captain says I don’t have to. Other than that, I’m as comfortable as possible while flying. I know that, on balance, air travel is a well-regulated system staffed by highly trained crews with layers and layers of safety precautions and a dedication to learning from past accidents. Let’s keep it that way.
Zeynep Tufekci writes a column for the New York Times.
Related Articles
Karl W. Smith: Raising Social Security’s retirement age is the least bad option
Ludwig, Miller: We are seeing a lethal shift in America’s shooting crisis
Allison Schrager: What are the odds of enjoying March Madness now?
Lisa Jarvis: Parents alone can’t protect kids from social media
David Brooks: The great struggle for liberalism