Foreign Airline Safety versus U.S. Major Airlines

by Philip Greenspun, ATP, CFI; December 2009, updated July 2011

Site Home : Flying : One Article


Malcolm Gladwell's business bestseller Outliers: The Story of Success has a chapter devoted to explaining the superior safety record of American major airlines compared to foreign carriers: "The Ethnic Theory of Plane Crashes". Gladwell comes to the conclusion that foreigners are unsafe because they are ... foreign. They have a strange and defective culture that prevents the first officer (copilot) from speaking up and pointing out problems to the captain. If only everyone were American, the world would be a better and safer place.

This article explores an alternative explanation: foreign airlines do comparatively poorly because their first officers have almost no pilot-in-command experience.

Pilot in Command

A soloing student is pilot in command. He or she may have consulted with an instructor before planning the flight, but once airborne the student has the final authority as to whether to continue the flight.

The holder of a "Private" pilot certificate is pilot in command when taking friends from San Jose's Reed-Hillview Airport up to Jonesy's Famous Steak House at the Napa Airport for lunch. He or she checks the weather, checks the airplane, and decides whether the risks are a reasonable match for his or her skills and equipment.

A Certificated Flight Instructor (CFI) is almost always pilot in command whenever flying with a student. The student is controlling the attitude of the aircraft, but the CFI is the one who decides when and where it is safe to fly. The CFI is the person who decides when it is necessary to take the controls from the student. The CFI is the person who decides that a thunderstorm is getting too close to the airport and it is time to cease practicing touch-and-goes.

Nearly all airliners require two pilots. The aircraft is too fast and complex for one person to operate safely. Both pilots are trained to be able to handle the airplane, run checklists, and communicate with Air Traffic Control (ATC). The pilots tend to switch flying/monitoring roles on every leg. But one pilot is designated "Captain" and one "First Officer". The captain is the pilot in command and has the final say as to whether to launch, whether fuel and weather make it prudent to divert to an airport other than the planned destination, etc.

Importance of the CFI step

A flight instructor is the only kind of pilot who, on a regular basis, sees aircraft being flown badly and needs to decide into which of the following categories the badness falls: (1) sloppiness that has no safety implications, (2) a potentially unsafe situation that can be ignored for a period of time until the student notices it and self-corrects, (3) an unsafe situation that must be corrected within 10 seconds by giving explicit instructions to the student, and (4) an unsafe situation that must be corrected immediately by the instructor assuming the controls.

This is very valuable experience in all kinds of commercial flying. A captain needs to judge whether a first officer's landing is going to be good experience for him, despite using up some runway and jostling passengers, or whether the landing is going to result in bent metal or running off the end of the runway. A first officer needs to recognize when the captain is burnt out from a 4-day trip and beginning to make mistakes.

Foreign versus U.S. Major

Buttonhole any pilot in a U.S. commercial airport and you'll learn that the major airlines hire only those pilots who have previously been captains of regional airliners or military planes. And the regional airlines, which supply most of the nation's major airline pilots, mostly hire from among those who have been flight instructors for 750-1500 hours.

A foreign major airline, by contrast, does not have a large pool of regional airline pilots, ex-military folks, and flight instructors from which to draw. Most foreign countries do not have an infrastructure of airports, flight schools, and private pilots. There would be no work for a flight instructor in such a country. Unless the country is very large, there won't be any regional airlines. Due to the shortage of qualified nationals, the foreign airline may screen young people and send the most promising to flight schools in the U.S. until they are trained to the minimum legal standards. For example, Japan Airlines runs a training center in Napa, California. Lufthansa trains its pilots in Arizona. A 23-year-old who can barely speak English and barely knows how to fly can go directly to the right seat of an Airbus.

How do those young Lufthansa pilots do? In March 2008, the 24-year-old pilot of a Lufthansa Airbus A320 attempted to land in Hamburg in a 50-knot gusty crosswind. The crew rejected a more favorably oriented runway at the same airport. The left wing contacted the ground during the landing attempt, ripping off the winglet. The 39-year-old captain took the controls and, with 131 passengers in the back, became a test pilot, adding power while the plane was still on the runway and taking off again (something easily done in a little Cessna, but not typically approved in a transport category airplane due to the fact that the engines take 7-9 seconds to "spool up"). "24-year-old co-pilot made first landing attempt" (Der Spiegel) has a video of the operation.

Would a 24-year-old American, equipped with a different ethnicity and culture, have done better handling the Airbus A320? There is no way to know because no U.S. airline would ever allow a 24-year-old to get near the controls of a full-size airliner. The first officer (copilot) of a U.S. major airline would be a 39-year-old with thousands of hours of experience as a captain of a 50-seat regional airliner. The captain of that airliner would be a 55-year-old who had been flying for three decades.

Let's look at typical foreign and domestic career paths side by side, with true pilot-in-command time highlighted in red:

Foreign Airline Pilot U.S. Major Airline Pilot
Private Certificate Obtained at professional flight school, includes 10 hours of solo. Obtained at flight school closest to house, including 10 hours of solo.
Poking around flying visually none 50 hours solo or with non-pilot passengers.
Instrument Rating 50 hours with an instructor. 35 hours with an instructor; 15 hours as P.I.C. with a friend acting as safety pilot.
Commercial Certificate 50 hours with an instructor. 10 hours solo. 20 hours with an instructor; 50 hours solo and building time taking friends and family for rides.
Multi-engine Rating 20 hours with an instructor. 20 hours with an instructor.
Flight Instructor Training ---- 20 hours with instructor; 10 hours practice teaching of friends
Flight Instructor ---- 1000 hours with students
Airline Training ---- 50 hours in a simulator at a regional airline
First Officer at Regional Airline N/A after roughly 1250 hours of flying, including more than 1000 pilot-in-command hours
Captain at Regional Airline N/A after roughly 3000 hours and four years of flying as first officer; an additional 2000 hours
Major Airline Training 50 hours in a simulator. after roughly 2000 hours as captain at regional airline, 50 hours in a simulator
First Officer at Major Airline after 250 hours of flying and perhaps 20 hours of solo after more than 3000 hours of pilot-in-command time, an additional 1000-5000 hours
Captain at Major Airline after about 20 hours of pilot-in-command time after more than 3000 hours of pilot-in-command time and perhaps 9000 hours total time

The table shows that the typical first officer at a major U.S. airline will start with more than 3000 hours of pilot-in-command time, i.e., roughly 150 times the P.I.C. experience of the first officer at a foreign carrier. The new foreign captain has essentially no experience as pilot in command. His or her P.I.C. time consists of 20 solo hours in a little Cessna on sunny days in Arizona or California. The new domestic major airline captain already has thousands of hours of experience as a captain; he or she was captain at a regional airline, doing exactly the same job but with a jet holding fewer passengers.

[A popular misconception is that most U.S. airline pilots are former military pilots. This doesn't make sense when you think about the tens of thousands of airliners that are up in the sky 12+ hours each day and compare to the comparative handful of military planes, each of which might fly only one hour per day. The military does not train nearly enough pilots to supply the airlines and, in any case, a retired military officer may not want to take a low-paid job that requires being away from his or her family 15-20 nights per month. That said, a retired military pilot can typically skip the regional airline step and go straight to the first officer job at a major airline, upgrading to captain when and if seniority allows.]

U.S. Regional Airlines

If having a newbie in the right seat represents a challenge to safety, how do U.S. regional airlines manage? The U.S. regional airline is hiring pilots who are more experienced than those hired by foreign airlines (the U.S. regional F/O typically has been a flight instructor) but less experienced than those hired by a U.S. major airline. The answer is that U.S. regional airlines have an accident rate as much as 5-10 times that of U.S. major airlines. The culture is the same, but apparently there is no substitute for experience.

Making all U.S. Airlines Safer

We're safer than foreigners, though not because of the ethnic or cultural superiority that Malcolm Gladwell has discovered. Could we be yet safer without making air travel much more expensive? Here are two simple suggestions: (1) give all pilots at an airline roughly the same schedule, and (2) provide good resting/sleeping rooms at commercial airport terminals for airline pilots who are between flights.

U.S. airlines are nearly all unionized. Union contracts are negotiated by very senior employees. It should come as no surprise therefore that a senior pilot might be able to arrange his schedule so that he need only work 8 or 10 days per month. He will be able to choose his home base so that it is close to his actual house. The senior pilot will have a short commute and up to three weeks per month of rest. How can the airline staff its flights if senior pilots are working so few days? By working the junior pilots 22 days per month, 16 hours per day (hotel room to hotel room).

Can a junior pilot legally fly for 16 hours in a row? No. Nor does he or she get paid to work all 16 hours; the airline only pays pilots from when an airplane leaves the gate until it arrives at the destination gate. A typical 16-hour day may include a 6-hour stop at an airport where the airline does not have a base and therefore there will be nowhere for the pilot to rest. He or she will be sitting near a gate, in uniform, reading a book, trying to shut out the noise of thousands of passengers walking by and hundreds of public address announcements. This kind of schedule is endured both by pilots newly hired by the airline and by pilots who have recently upgraded to captain. A passenger has no way to know if the crew is a well-rested senior crew with a "princess" schedule or an exhausted junior crew who are "on reserve" (essentially working right up to the legal/human limits).

How does it work for pilots of private airplanes? Netjets is the leading business jet operator. Their unionized pilot workforce shares the same schedule for everyone: 7 days on; 7 days off. When there is a 6-hour stop between flights, even the smallest airport (what most people would call an "airstrip") will have sofas, recliner chairs, and sometimes even small bedrooms for naps. The pilot of a $30,000 four-seat single-engine prop-driven airplane has a better place to rest than the crew of a 300-passenger jet.

To improve safety at U.S. airlines, the seniority-based scheduling system should be eliminated and commercial airports should be required to use the fees that they collect for each passenger to build a reasonably comfortable quiet lounge for crew rest.

More: see the comments on a May 21, 2009 blog posting on the same subject.

[Note that the work-the-junior-guys-to-death systems is not universal; quite a few foreign airlines already schedule crews without regard to seniority.]

Could the "Ethnic Theory" be Correct?

It is certainly possible to find some flight instructors who believe that a certain ethnic group or nationality is more or less suited to flying. Can we use this anecdotal evidence to support Gladwell's "Ethnic Theory" of flight safety? Gladwell supports a few aviation anecdotes by reporting on some research on "power distance index", the likelihood that a junior person in a hierarchy would feel comfortable speaking up to a senior person. One brief study was even done on pilots in various countries. But on balance a generalization about a culture may not apply to pilots within that culture. The U.S. Navy did a psychological study of its female pilots and found that they had more in common with male Navy pilots than with average American women.

Ethnic and cultural factors are difficult to measure. Did the 24-year-old Lufthansa pilot not speak up about the inadvisability of attempting a landing in a crosswind stronger than the airplane's maximum demonstrated capability because she was a woman? Because she had been reared in Germany? Does a Korean first officer not speak up because he is Korean? How can we possibly know? In the areas that we can measure, such as years of flying experience and hours of pilot-in-command time, these pilots are not remotely comparable to the first officers of mainline United Airlines, Southwest, JetBlue, Fedex, UPS, etc. So why are we comparing their performance?

Conclusion

Gladwell makes a complex case about how the inferior culture of foreigners leads to inferior safety. A much simpler explanation for any safety differences is the difference in experience:

About the Author

Philip Greenspun holds FAA Airline Transport Pilot and Flight Instructor Certificates with ratings for multi-engine airplanes and helicopters. Greenspun has type-specific training for the 50-passenger Canadair Regional Jet (CRJ) and the Cessna Mustang business jet. Greenspun has flown as first officer of the CRJ in scheduled airline service from bases at JFK and Cincinnati to all of the major airports of the Northeast.

About Little, Brown

"The Ethnic Theory of Plane Crashes" was published by Little, Brown as part of Outliers. At the $28 list price, a consumer might expect that some fact-checking had been done. Yet all of the information on this page could have been learned by a Little, Brown intern who stopped by any major airport in the U.S. and talked to a couple of pilots. Of the information presented in the chapter, much is plainly incorrect and would have been red-penned if the chapter had been submitted to any airline pilot for review. For example, on page 214, a Korean Air first officer "got confused while listening to Air Traffic Control [ATC] and mistakenly put the plane on a course intended for another plane. ... Finally the plane's radar picks up the mistake." A pilot reviewing the chapter would have noted that ATC probably assigned a heading (direction to point the nose of the airplane) rather than a course (track over the ground, which will depend on the speed and direction of winds aloft and is not easily controlled by pilots of older aircraft). More importantly, a reviewing pilot would have noted that a jet's radar cannot "pick up a [heading] mistake". Airborne radar is also called "weather radar" and is used to detect precipitation in front of an aircraft. The radar signals are sent out from the nose of the plane and reflect off the back of raindrops. Heavy reflections indicate heavy rain, which is important because in most parts of the world, heavy rain is associated with thunderstorms, which are extremely turbulent and hazardous. An airplane's radar cannot see other airplanes nor provide any information about heading or course.

A Sri Lankan pilot is quoted as saying that New York air traffic controllers are "rude, aggressive, and bullying" and that this may have contributed to a Colombian Boeing 707 running out of fuel and crashing (full story). A Little, Brown intern could have listened to JFK Tower (aircraft from the ground up to about 2,500') and New York Approach (aircraft from 2,500' up to 10,000') on liveatc.net and checked for rudeness or "bullying". I have flown four-seat airplanes, four-seat helicopters, and 50-seat airliners in and around New York City. Typical for me was an instrument flight rules (IFR) trip from Boston to Washington in a Cirrus SR20. A 50-knot headwind right on the nose was going to make me late for my arrival at the Gaithersburg, Maryland airport, where my parents had planned to meet me. The routing was right over the top of JFK at 6,000'. I broke into the busy New York Approach frequency to ask for a temporary radio change to Flight Service, the guys who handle flight plans, weather briefings, and other non-urgent matters. After vectoring a few 747s heading out to Europe, the New York Approach controller asked me what I needed. I explained that I was hoping to convince Flight Service to call my mother in D.C. and advise her of my new arrival time. "We can do that for you, buddy," said the controller. He took down my phone number and had his assistant call my mom.

[What if the controller won't call your mom and/or is otherwise unhelpful? One of the first things that you learn during flight training is that the pilot in command can declare an emergency and, after that, do whatever he or she deems necessary for safety, regardless of the regulations or ATC instructions. The pilot can simply say to ATC "Avianca 52. Emergency. Low fuel. Starting the ILS 4R approach." It then becomes ATC's job to clear other aircraft out of the way.]

No "Malcolm Gladwell" is listed in the online FAA Airmen Certification registry, from which one can infer that he has no flight training or pilot experience. His biography says that he majored in history, from which one can infer that he has no technical education. Yet apparently his publisher, Little, Brown, decided to publish a chapter on how to fly airliners without asking a few pilots to review it. Then they charged consumers $28 for the unchecked manuscript. Can someone please remind me how traditionally published material is supposed to be more reliable than Wikipedia?


Text and photos Copyright 2009 Philip Greenspun.
philg@mit.edu

Reader's Comments

Malcolm Gladwell made the mistake of ascribing a very complex set of circumstances to a single cause. You made the same mistake. In reality, culture, experience, and a host of other factors all come together to determine mishap rates: maintenance quality, equipment age, safety equipment, company procedures, ATC quality, airfield quality, predominant weather conditions, terrain, migratory bird patterns, language barriers, etc. Experience is almost certainly a bigger factor than culture, but culture is still a very big factor: why else would the majors spend millions of dollars every year on CRM training? To see just how much of a factor this is, look at how mishap rates plummeted after the introduction of CRM training at US carriers in the late 1970s/early 1980s. Prior to widespread CRM training, the culture at major US carriers was largely as Gladwell describes current foreign carrier culture: the captain's word was final, and co-pilots didn't speak up or contradict the captain. Multiple studies found this to be a major source of mishaps, which is one of the major reasons for the introduction of CRM training. You also have to consider how inexperience is exacerbated by this type of culture: consider two equally inexperienced co-pilots who each notice the same unsafe condition on two separate flights. The co-pilot for the airline with bad culture will be afraid to point out the problem, thinking to himself, "My captain has thousands of hours more experience than I: he certainly must have noticed the problem, and would do something about it if it concerned him. I don't want to incur his wrath and/or ridicule by distracting him with such a minor problem." The co-pilot for the airline with good culture will be far more likely to speak up, possibly averting a mishap.

-- Lucas Jung, December 26, 2009
Lucas: CRM is well-known and has been taught for decades at foreign carriers as well. As far as the improving safety record of U.S. airlines goes, consider that the period you're describing is also the period when terrain awareness systems were installed (as far as I know, no airliner equipped with a moving-map GPS-based TAWS system has been crashed into a mountain), when engines became much more reliable, when systems in general became simpler and more reliable, and when ground-based systems were substantially enhanced (weather radar feeds to ATC, more ILSes, etc.). The handful of hours of CRM training delivered to airline pilots during those decades may have made a difference, but (1) the effect is hard to separate from safety improvements due to improved technology, and (2) CRM training was delivered more or less equally at foreign carriers.

-- Philip Greenspun, December 29, 2009
As a 20 year airline pilot I do disagree with the US vs foreign pilot argument, but I will leave that alone as it will involve pages of refutory evidence. There is some basis of fact in what you say, but it is simplistic and, well, thats enough on that...

As for flying to JFK. New York is to most airlines who fly heavy jets effectively a third world airport. It is a nightmare to taxi around, as it is way way to small (was designed for 707's max) and aside from seemingly nothing working the way it should and being way way too overcrowded, the controllers make no allowance for non US-English speaking crews. My airline flies around the world and the US is difficult full stop. Unlike almost anywhere else it takes a LOT of practice and understanding of US culture to fully know what is really being said. And our airline is also from an English speaking nation. The problem is that US pilots and controllers _generally_ only operate in the USA and deal with people who have grown up in the same country. For example, at LAX, a standard call on GRD frequency was to switch from Nth to Sth controllers 'at the 50 yard line'. What the HELL does that mean for anyone who did not happen to play US college football in the USA?? Another example - even in China you might be cleared direct to a VOR. They will NEVER make a comment (in heavily accented and highly colloqial US English) like 'JA you can head to Urumqi when you can'. They will speak in concise and aviation standard 'AJA when clear of weather direct URC'. And Americans make it way worse. URC VOR can be reasonably deduced as being related to Urumqi. What about something like the Wilkes Barre VOR's ident of LVZ?? Yet US controllers NEVER use the LVZ terminology, but always the full (and often indecipherable) 'JA you can head to Wilkes Barre when you can'. Totally foreign words to totally foreign ears about totally foreign places. Do you turn left or right? How do you spell Wilkes Barre? Which Map is it on? Is it an FMC Ident or a place name? For those reading this who don't know, aviation charts are not labelled clearly with place names, but three letter code Morse code identifier letters. And at JFK in particular, the talent for running seemingly dozens of taxi instructions at you in one sentence of very heavy New York 'English' is second to none. Sure the controller may say the same thing 200 times a week for 3 years, but to a foreign crew (who might only operate into JFK once a year) and after being at work for sometimes up to 20 hours at a time, such gibberish is at best near indecipherable, at worst dangerous with just the aid of a dimly lit small chart on possibly slippery, icy and the certainly very narrow taxiway's of JFK. And in heavy snow, the taxiway lights at JFK are non existant. I heard one Asian crew make a very understandable error (so understandable we earlier had to stop and get abused by the controller to make sure we didn't make the same exact error). The extremely irate controller said extremely quickly in a VERY thick NYC accent 'XYZ if ya cant follow simple instructions then we'll sin bin ya on Zulu (taxiway)' I sympathise with the workload at JFK, but it is absolutely NOT a favourable aviation environment for anyone without a VERY thorough understanding of how things 'really' work there. We once got stuck along with everyone else in a bizarre traffic jam in JFK. In a beautiful crisp English accented voice the words came over the radio 'I am slowly losing the will to live (at JFK!)' Americans CAN do it very well. It took a while for them to really standardise the operation to what the international rule book says and also for them to understand 'international' English. The internatinal aviation rule book which, it must be said, was worded primarily by the US so you would think it should be fairly easy to get it right... However, once they did master it, the controlling the USAF now provides over Afghanistan in particular (which used to be horrendous with minimal equipment and multiple countries who all hate each other converging in one place etc etc) is now exceptional. I am sure I am not alone in the hope those controllers go home and mix with their US couterparts and explain that it is often very hard to operate over foreign countries with many different accents and backgrounds and work out what it is that people (pilots) are really trying to say to each other and to Air Traffic Control. Familiarity is great (and often very humorous) but only if you are actually familiar!!!

-- james lawson, May 10, 2010

The US is a little over 4% of the world's population. Can you see how parochial it sounds to speak of the 96% of the world as a single group, inferior in this case, and label them "foreigners". There is a huge range of safety performance of airlines across the globe, including in the USA. I don't feel as safe on an American airline as I do on some Asian, European and my own Australian Qantas. The whole analysis comes across as being in a Xenophobic spirit. The original author's thesis would not have been out of place in the intellectual milieu of master race superiority that lead to Nazi Germany,

-- Ashley Bear, June 9, 2010
I think you have rather substantially misinterpreted the point of Gladwell's chapter in Outliers. It's not at all about the relative safety of U.S. versus foreign pilots. It simply picks up on the well known psychological literature looking at how cultural differences in "power distance" affect pilot performance. He really only singles out Colombia and South Korea--and in the case of South Korea points out how they overcame that cultural disadvantage. It's probably a good idea to read a book closely before being so scornful of its message.

-- Arthur Simons, September 4, 2010
Add a comment | Add a link