Friday, July 12, 2013

Are Korean airlines unsafe? Ask a Korean

The "ethnic theory of plane crashes" was first advanced by Malcolm Gladwell in his book "Outliers."  Gladwell's analysis is ably deconstructed by TK, the author of a blog called "Ask a Korean."  The piece is long, but it can be succinctly summarized as: Gladwell is full of shit.  And I mostly agree.

So does that mean that I am retracting yesterday's advice to not fly on Korean airlines?  No, I am not.  It is a fallacy to conclude that because someone advances a bogus argument for a position that the position is therefore wrong, even if the person advancing the bogus argument is a famous best-selling author.  In my judgement as both a pilot and someone with scientific training, Korean airlines are less safe than other airlines.  The relationship is causal.  It is cultural (not ethnic!).  It is a result of systemic flaws in Korea's training regimen for pilots.  I base these conclusions not on Gladwell's analysis, which is indeed deeply flawed, but on my own analysis of primary sources, specifically, the historic track record of Korea's airlines, the information currently available about Asiana 214, and the anonymous piece written by someone claiming to have first hand knowledge of the current situation in Korea's training program.

One should, of course, be extremely leery of anything anonymous one finds on the internet.  However, in this case I believe that this piece is genuine, that is, it was written by someone who is what they claim to be.  Whoever wrote it is clearly a pilot, or at least intimately familiar with flying terminology and culture.  It is very hard for a non-pilot to fake being a pilot.  There are subtle uses of language that almost always give a faker away.  My favorite current example: TV reporters who call the runway that Asiana 214 landed on "twenty-eight left."  No pilot would ever say that.  It's two-eight left.  The anonymous piece throws around a lot of lingo but contains no such errors.  So if it's a fake, someone would have had to go to an awful lot of trouble to fake it.  And to what end?  I see no plausible motive for forging a piece like that, particularly since the original was deleted.  All the data is consistent with the theory that the piece is what it appears to be, a genuine first-hand description of deep systemic and current problems in the Korean airlines training system written by someone who experienced them first hand and was frustrated in their attempts to try do something about it.

However, I will temper my position with one observation: notwithstanding the problems in Korea's pilot training system, and notwithstanding that Asiana 214 crashed because (based on the currently available information) of basic pilot error under the least demanding conditions possible (visual landing, no wind, perfectly operating state-of-the-art aircraft), flying on a Korean airliner is still fairly safe relative to other risks that people routinely take (like driving to the airport).  So if you have to get somewhere and KAL or Asiana are your only options, sure, go for it.  But all else being equal, I would avoid these airlines until they get their shit together.

UPDATE: The Christian Science Monitor has published a story about this.


Don Geddis said...

I read Outliers, and enjoyed it. I remember the plane crash chapter. Your "Ask a Korean" reference has an excellent takedown. Thanks for the link.

Miron Brezuleanu said...

20 years before "Outliers", Donald Norman mentioned a Korean plane disaster in "The Design of Everyday Things". I don't remember the flight number, it was about a Korean passenger plane shot down by the USSR because it strayed into their airspace. The root of the problem was a design issue in the flight computer, but the actual cause for the crash was social pressure (the pilots were pressured to work with a defective flight path rather than follow the safety procedure and return to the airport).

Granted, this was a two page discussion at most, not a chapter :-)