Skies Are Actually Significantly Friendlier


President Trump’s ridiculous tweet this morning – taking credit for a record low number of civil aviation fatalities this year – made me think about the remarkable record over the last two decades.

This matters to everyone at some level. But it has a particular relevance to me as a semi-recovered aerophobe. Statistically speaking, commercial air flight in the US has always been very, very safe. You know all the statistics about how you’re far more likely to get killed in a car crash on the way to the airport than flying on the plane. (Or maybe that’s just me? They tell you this stuff a lot when you have a hard time getting on a plane.) But it is amazing how air traffic safety has changed over the last two or three decades.

I’m about to turn 49. If you’re my age or nearabouts you remember that in the 70s and 80s it was pretty much expected that every year there’d be at least one big jet that would go down killing nearly everyone on board. Even since the early 80s the numbers are daunting.

1982: Air Florida – 70 dead
1982: Pan Am – 137 dead
1985: Delta – 126 dead
1985: Arrow Airways – 248 dead
1987: Northwest – 148 dead
1988: Pan Am – 243 dead
1989: Independent Air – 137 dead
1989: United Airlines – 110 dead
1994: USAIR – 127 dead
1995: American – 152 dead
1996: Valujet – 105 dead
1996: TWA – 212 dead

Now these aren’t all the crashes. With the exception of the Air Florida crash, a shocking and largely televised crash in which a plane went down into the then-icy waters of the Potomac River right outside DC, I’ve only included disasters with more than one hundred fatalities.

We of course had four big terrorist-driven air crashes on 9/11 – notably none of the four planes had more than 100 people on board. Then just two months later an American Airlines flight went down in Belle Harbor, New York. That’s a crash right after takeoff from JFK into a neighborhood in Queens. Given that it came so soon after 9/11 it at first seemed very hard to believe this wasn’t another terrorist attack. But it wasn’t. It was mix of pilot error and an underlying design flaw.

But that was the last major plane disaster in the United States or with American carrier. More than 16 years ago. That doesn’t mean no one has been killed in a commercial plane crash. Four commuter jets had crashed since 2001. Those have historically been less safe than the commercial jetliners the major airlines fly and generally carry many fewer people on board. Commuter planes went down in 2003, 2005, 2006 and 2009. Since 2009 there’s been no major air disaster in the US at all.

At first this just seemed like a lucky streak. But it’s been going on long enough now and the difference from the past is so dramatic that it’s clearly no coincidence. I’ve done a little research on this. Knowledgable people think the most important change may be something called ‘Crew Resource Management‘. I had to look up again what the training was called. But the basic gist is a different model of teamwork and communication that flight crews use during emergencies. It’s complex but also deceptively simple. But one key point is a somewhat less authoritarian command model. The captain is still in charge. But junior officers (first officers, flight engineers, etc.) are encouraged to say something if they think the captain is doing something wrong or missing something. See something? Say something. (Just now I found this 1987 article from The New York Times which discusses what I think were the origins of the movement.)

There’s a whole complex story about how this model came into being. The principles themselves sound pretty elementary. But they emerged from a realization that most flight disasters occurred because of relatively minor (or at least not necessarily catastrophic) mechanical problems that triggered a series of pilot errors leading to disaster. More effective teamwork, more input and communication from everyone involved, getting everyone’s input, could make a big difference.

To be clear, I’m no flight safety expert and don’t want to present myself as such. I’m not trying or in any position to say the increasing safety of US air travel is all or mostly do to CRM. I’m sure better planes, often made better by learning from earlier disasters, and more experience running the country’s mammoth civilian air flight system play key roles. I simply mention this because this subtly different way of training flight crews to work together is seen as a major part of the story.

I should also note that air flight was of course very safe in the 1970s too. It’s just even safer now. And if you’re an aerophobe, knowing that the last time a big jet went down was almost 20 years ago makes getting on the plane and those moments of turbulence a bit easier, even if the purely statistical difference is not that great.