Germanwings Tragedy: How to protect against mentally ill pilots?

When friends first asked how the Germanwings plane could have descended out of cruise flight and crashed into a mountain on a clear day, I responded with “EgyptAir 990“. Sadly it seems from the information that has been released this is indeed what happened.

Friends are now inundating me with questions about airliners, cockpit security, and how to protect the plane and passengers against a mentally ill pilot. Here are some examples:

Why don’t airliners transmit flight data recorder (“black box”) and cockpit voice recorder data to ground stations that so we get quicker answers to what went wrong? Partly this is due to a lack of a high-bandwidth worldwide network through which to send the data. The Iridium system is old and slow. The usual suspects of Silicon Valley world-savers are supposed to come through with a 700-satellite constellation by 2019 (OneWeb). Perhaps at that point it would be practical to send a wide stream of data out but airlines would no doubt resist the fees for FAA-certified hardware ($1 million per airplane?) and data charges.

Why does TSA screen pilots so thoroughly before they get on planes? I don’t know. As a regional jet crew of three we used to have to slog our way through security every morning in our airline uniforms with our forgery-proof IDs. TSA would hassle us thoroughly if we had a few extra coins in a forgotten pocket. Yet as soon as I sat down in my seat I could easily reach for a truly nasty weapon: the crash axe. It was required equipment on all of our planes, weighed about 5 lbs., and looked like something a medieval warrior on an HBO show would wield. We had no training regarding how to use this device and I am not sure what it is for (breaking the windows to escape the cockpit? we had an escape hatch on top). However, I am confident that it would make a better weapon than two quarters and a dime, a set of car keys, or a mobile phone (items assiduously sought by TSA).

Could this crash have been prevented if a flight attendant had been in the cockpit? “No,” is the short answer. The EgyptAir 990 captain managed to get back into the cockpit fairly quickly and still could not save the plane. It turns out that things can go bad very quickly in a jet. Pushing one button will cut off the fuel to an engine (i.e., two presses on two buttons and a standard twin-engine plane will no longer have power). Stomping on the rudder can break off the tail of an Airbus (American Airlines 587). Pressing another button will dump the cabin (instant depressurization). Pushing forward on the stick or yoke can put the airplane into a 30-degree nose-down attitude within a few seconds. The airplane can be rolled at 30 degrees per second until it is upside down. Automated systems can be disabled by pulling circuit breakers. Imagine that Pilot 1 gets up to go to the bathroom and the flight attendant has just come into the cockpit and locked the door. Pilot 2 initiates the preceding actions before the flight attendant can get seated and belted in. What chance does Pilot 1 or the flight attendant have to save the airplane at this point?

[Separately, remember that it is also fairly easy to crash a jet at low altitude through simple incompetence. The mentally healthy pilot probably wouldn’t have time to react to things done by a mentally unhealthy pilot if done shortly after take-off (plane is going 180 mph and is full of fuel at that point). You thought Captain Sully was a hero? Remember that the Pilot 2 can create that situation for Pilot 1 by pressing two buttons. The first officer couldn’t fix the captain’s fairly simple mistake in time on Colgan 3407.]

Were you subject to pre-employment mental health screening? “Not really,” is the best answer. Delta mainline pilots are, supposedly, with the MMPI and a psychologist interview. But we just got some tests for “cognitive skills” (ability to multi-task and avoid distractions and/or concentrating too much on one thing) and a decision-making test that was vaguely psychological. Perhaps the company was afraid to see just how crazy pilots had to be to take a 22-day/month, 16-hour/day job for $19,000 per year.

Could a crazy or depressed pilot get released to fly with passengers? Probably not. Training takes three months and a lot of people have an opportunity to interact with you in what can be intimate surroundings. There are a couple of checkrides with FAA-designated examiners (previous posting on the subject of training). There is a stage called “initial operating experience” with “check airmen” (sexist nomenclature but some are women) of 50-120 hours in a real airplane with real passengers and people who are truly unsuited to the job often wash out at this point by failing to be signed off. (Obviously there is some risk of a pilot intentionally crashing the plane during this phase, though the check airmen are presumably on high alert.)

Could a crazy or depressed pilot continue to fly with passengers? Yes. Due to the fact that it is illegal for a foreign airline to come in and put all of the U.S. carriers out of business (Ryanair’s costs are only about 45 percent of Southwest’s (source)), U.S. airlines are fat targets for labor unions (JetBlue pilots unionized recently). It is a lot easier to fire a pilot than a public schoolteacher, but not simply because the pilot doesn’t smile and fails to respond to “How are you doing?” with the standard crew lounge response of “Livin’ the dream!”

Would it help to have annual psychological testing? It would certainly help the psychology industry! One issue is that psychological tests were designed for populations of people with serious mental illnesses. They may be meaningless in a population of more or less sane people. Academic studies have shown that antidepressant drugs may not be more effective than placebos and that therapy may not be any more effective than drugs (example). If psychologists can’t treat depression consistently how do we know that they can recognize it? I flew with some captains who had quirky personalities. For example, while acting as “pilot monitoring” on one flight it was my job to press all of the buttons while the “pilot flying” was hand-flying the aircraft (one hand on yoke, one on the thrust levers). The plane was on autopilot and we were assigned a new altitude by air traffic control (ATC). I started pressing some buttons that now properly belong to the captain, a company check airman. “Who’s fucking this sheep?” he asked. What would a psychologist have done with that? And what would a psychologist do with a population of notorious malingerers who share information on what kinds of questions are asked and what the right answers are? There are entire web sites devoted to the topic of how to game the airline interview process. Due to the industry’s unique characteristics (see “Unions and Airlines”), the consequence of being fired from a $250,000/year job (captain with seniority at Airline A) will likely be starting a $50,000/year job doing the same tasks 2X as many hours (first officer with no seniority at Airline B). So a unionized pilot would have far more incentive to try to beat a psychology evaluation than would an ordinary worker.

How does pilot mental health compare to the average? Pilots need to pass a medical exam every year. If they exhibit serious mental health issues the doctor probably wouldn’t issue a medical certificate. Learning to fly requires a certain amount of mental discipline. Commercial pilots are subject to random drug testing so it is impossible to be on drugs or alcohol much of each month. Thus my layperson’s perspective is that pilots tend to be reliable, patient, and sober/drug-free. There is one big exception, however: divorce, custody, and child support. Pilots are away from home 10-22 days/month. Suppose the stay-at-home spouse decides he or she is bored and needs to have a lover. If the stay-at-home spouse progresses to the plaintiff stage, most U.S. states will reward that spouse with the house, the children, and at least half of the pilot’s income going forward. The cash and the house go with the kids. The pilot is the slam-dunk loser for any custody lawsuit because he or she was away much of the month and therefore cannot meet the “historical primary caregiver” standard that is used when courts allocate children and the child support profits that accompany them. “Suing a pilot is almost as easy as suing someone deployed overseas in the military,” is how one litigator put it. What if it is the pilot who indulges in sexual activities on some of those nights away from home? Here’s an excerpt from our Massachusetts chapter:

“There are a lot of women collecting child support from more than one man,” Nissenbaum noted. “I remember one enterprising young lady who worked as a waitress at Boston’s Logan airport. She targeted three airline pilots, had a child by each of them, and back then was collecting $25,000 in tax-free child support from each pilot. Of course, instead of serving food and beverages, she did have to care for those children.” Using the USDA-estimated actual costs of children and the 2013 Massachusetts child support guidelines, compared to the college/work alternative a woman would have more spending power collecting child support from two men when each had an income of at least $95,000 per year (sufficient to generate $39,264 per year in tax-free income or $377 per week from each defendant).

Pilots get sued often enough that people talk about AIDS: Aviation-Induced Divorce Syndrome. And with our winner-take-all divorce system in the U.S., the loser will tend to be unhappy about it. Here’s an excerpt from our Children, Mothers, and Fathers chapter:

Attorneys we interviewed reported that some of the male defendants that they’d sued had killed themselves. One lawyer noted “The guy never would have signed up to be a secondary parent but now that is his court-ordered role, along with paying all of the bills for a mother who was greedy enough to turn her child into cash. Most of the guys who are worth suing are middle-aged. You might be talking about a 50-year-old defendant with a 1-year-old child. He’ll be paying until he is 70, but the amount that he pays was set when he was at his peak earning capacity. As he gets older he’ll probably be earning less, but judges almost never adjust child support or alimony downward. So gradually he’s paying an increasing share of his income to his plaintiff. I’m actually surprised that more of these guys don’t kill themselves. They really have no reason to exist other than to write checks to someone with whom they were briefly acquainted many years previously.” What do the researchers say? “Divorce and suicide risk” (Kposowa 2003; Journal of Epidemiology and Community Health 57:12): “Divorced men were over eight times more likely to commit suicide than divorced women. … After taking into account other factors that have been reported to contribute to suicide … [divorced men] were nearly 9.7 times more likely to kill themselves than comparable divorced women.” For comparison, “Marital Status and the Risk of Suicide” (Smith, et al 1988; American Journal of Public Health 78:1) shows that married men 45-54 kill themselves at about 2.5 times the rate of married women of the same age.

If a pilot is depressed because he (typically) lost a divorce, custody, or child support lawsuit, can’t he simply ground himself for a while? The answer is “yes” from the airline’s point of view. From a practical point of view the answer is “not” unless he wants to go to prison and never work again. Remember that if was worth suing he probably was earning towards the higher end of the pilot pay scale. There aren’t going to be any non-flying jobs with comparable pay. And he is depressed so who would hire him? Thus he has a child support and/or alimony order in place based on his $150,000/year pilot salary. If he can’t earn close to $150,000/year he won’t be able to pay this order, in which case he will join the roughly 1 in 7 child support payors who are imprisoned at some point for nonpayment (for not paying the court-ordered amount, Massachusetts offers the pilot a felony conviction, which means he won’t have an ATP certificate anymore and therefore won’t be able to work again, plus up to 10 years in prison). Thus the pilot is given a powerful incentive by the divorce court, his plaintiff, and all of the lawyers involved to keep flying the American public, regardless of how depressed or suicidal he is feeling.

What about fancy technology fixes? Given that the airliners can be flown by hand and that most of the automation can be turned off by pulling circuit breakers accessible to the pilot, I don’t think that there is a way for current airliners to be made pilot-proof. Only the “big hammer” of a fully automated airliner would protect passengers from pilots intentionally crashing airplanes. Are we ready for complete automation? At the pace of progress that is permitted in our regulatory environment I would say that it will take at least 15 more years. One potentially serious problem is that an airplane flown automatically is by definition being flown on instruments. Yet our air travel system collapses if planes have to make instrument approaches. Things only work at busy airports when pilots can make visual approaches (previous posting) and follow fairly closely behind other planes.

Some ideas

As noted in “Unions and Airlines,” pilots would be in a lot better shape if they all had a reasonable work schedule. So the FAA should require that airlines give all pilots equal schedules, regardless of seniority. A borderline suicidal pilot who hasn’t slept for five nights is worse than a borderline suicidal pilot who is well rested.

Airlines should be able to automatically fire divorce, custody, or child support lawsuit defendants. We have set up a system where the pilot is virtually certain to lose the house, the children, and most of his or her income going forward. After absorbing that kind of loss there is a pretty good chance that the pilot’s mental health will be seriously compromised. Our legal system has decided we no longer want that person as a parent for the children (except for a few days per month). Why then would we want that person having the power of life or death over hundreds of none-the-wiser ticket-buying customers?

TSA should stop hassling pilots. If nothing else, the Germanwings crash proves that a pilot doesn’t need accessories to endanger passengers. Maybe have a simple screening procedure for pilots and flight attendants to prevent them from bringing actual guns into the airport, but otherwise let them be. What would your mood be like if you had to be TSA-screened every working day of your life?

There should be some truly plush crew lounges at every airport where a pilot might have to wait. If I fly a little Cirrus to B. Coleman in Gary, Indiana I can exercise, enjoy a movie in a theater, work in peace and quiet with high-speed Internet, etc. If I’m an airline pilot with a four-hour break in some airport without a crew lounge I am stuck in the public terminal. That’s a depressing way to live even for a person whose mental health was great to begin with.

We should allow Mexican and Central American airlines to operate freely within the U.S. The Mexican suicide rate is much lower than the U.S. average (source). I’ve never met a Latin American crew member who didn’t seem to love his or her job (contrast to the Lufthansa pilots who have been out on strike 13 times in the last year (article)). We’d all be in a better mood if we could fly COPA domestically.

Closing Thoughts

One reason that the Germanwings and EgyptAir crashes are so horrifying to me is that ingrained in pilot culture is the idea that passenger safety and comfort are paramount. If we need to roast up in the cockpit so that they can be comfortable back in the cabin, the temperature knob is set accordingly. This was a sad week for aviation.

21 thoughts on “Germanwings Tragedy: How to protect against mentally ill pilots?

  1. Thanks for the insiders info 🙂 16 hour /day $19k a year looks scary given the amount of power a pilot operates.
    Anyway I hope this [crash] won’t result in hassling the pilots even more as a result. We are all prone to overreacting whenever illusion of 100% safety is broken.

  2. I think the focus on the pilot’s depression in the news now is somewhat misleading. What it looks like to me at the moment is a workplace spree killing like the Lockheed Martin shooting or one of the notorious U.S. Postal Service spree killings. At the end of these the disgruntled worker frequently commits suicide: in that light Lubitz’s decision to do something suicidal isn’t mysterious. Effectively he was just using his control of the aircraft instead of the more usual guns. Thwarted ambition, resentment towards the workplace, the ambition to finally make it big with a grandiose statement : the usual motivations of a workplace killer seem to be in place here. (I am not a psychologist, so this isn’t expert opinion.) IIRC these employees were usually thought of as disgruntled and unhappy, but usually not as obviously crazy or primed for mass murder, and in an unhappy workplace a resentful worker may not stand out much anyway.

    Which brings us to Germanwings and Lufthansa and those European airlines’ running costs. I know someone who travelled long-range on Lufthansa recently. From what he tells me, if Lufthansa had screened its staff for seething resentment towards the airline and its passengers then most of them would have come up positive. And a major cause of that is surely the old European flag-carrier airlines’ attempts to cut costs to compete with the budget carriers like Ryanair, and the predictable campaigns of resistance by their staff unions. One major tactic of the old airlines has been to create budget offshoots which operate like budget carriers, including of course reduced pay and conditions for pilots. And in fact it turns out that Germanwings pilots were on strike just last month !

    Finally, I really hope that the precautions against future incidents are effective, because IIRC these kinds of killings tend to attract copycats.

  3. The idea of firing divorced employees (I hope you meant it as a joke) is dangerous. By raising the stakes even more, the employee who is about to be fired may decide to kill not only himself but the spouse who has brought this upon him and maybe the kids and the judge and who knows who else.

    In a way it surprises me that MORE of those who are about to commit suicide don’t decide to go out in a blaze of glory, especially since we have guns everywhere. Thankfully, I suppose, the nature of depression means that most of the depressed only want to kill themselves.

  4. Apparently the demand of the strikers was that the current Lufthansa benefit which allows pilots to retire at age 55 at 60% salary should be applied to newly hired pilots as well. This sounds like a lovely benefit (however, 70% salary at age 50 sounds even better) – I would like something like that too, but since I am self employed I can’t go on strike to demand that from my “employer” and I have to keep working in order to get paid. I would also like free ice cream AND cake.

    Since Ryanair doesn’t have anything like that, it means that Lufthansa will go broke eventually if they give out benefits like this. Either they will match Ryanair on ticket prices and lose money on each fare or else they will lose passengers and keep shrinking. Either way is a death spiral.

    The TSA screening of pilots is just part of security theater. 90% of what TSA does is worthless and intended mostly to humiliate you. Where else do minimum wage losers get to order around rich professionals in our society? The frequent racial dimension adds another twist. Revenge is sweet.

  5. I suspect one reason for TSA inspection of flight crew is that IDs can be forged. If you give a pass to anybody entering the secure area, you change the problem from finding weapons to verifying ID. And create a big hole if you have any hole in your ID system.

    Clearly a suicidal pilot can crash the plane. But suicides are not like terrorists. They don’t just find another way if you block their preferred way. Sometimes if you block a suicide’s primary plan, you prevent the suicide. It’s entirely possible, for example, that if the captain had not left the cockpit on that 2 hour flight that the suicidal pilot would have done nothing that day, then gone home and reconsidered his depression or gotten more treatment, and never did a thing. And yes, had the captain been able to regain the cockpit and just talk to him, forget about taking the controls, it might have been different.

  6. Brad: Flight crews usually go through security together. Security experts have recommended dropping security inspection of intact crews where the pilots and flight attendants can vouch for the others. It is hard to forge an ID and also convince the captain and flight attendant that you were the person they flew with yesterday. (Why drop the focus on flight crews? It frees up resources to look at potentially more problematic passengers. Though if you model TSA as an agency with infinite access to dollars and staff this idea would not be helpful except to the mood of flight crews.)

  7. I agree, if they go through together you can have lighter security.

    But I also think we could do all the security in radically different ways, with bomb-proof luggage containers to reduce scrutiny of luggage, security appointments and more random screening. Plus understanding that once you’ve solved the problem of the bad guy getting control of the aircraft to use it as a weapon, the small jet is just another space like a train, bus, boat, stadium or airport security line where a guy with a bomb could kill a lot of people

  8. Though I will add that, if a bad guy can fake ID, why can’t a group forge a set of IDs for a group of flight crew? You need to require that the flight crew go through together and that they are also personally known to the screeners.

  9. >why can’t a group forge a set of IDs for a group of flight crew?

    This is the kind of scheme you might see in a Die Hard movie or something the Israeli’s might do, but garden variety terrorists are not evil geniuses, just evil and rarely do they have the skills and organization to pull something like this off. Much easier to just put explosives in your underwear or something like that.

  10. My late father-in-law was a pilot for the now defunct TWA. He worked during a time that had 2 pilots, and a flight engineer in the cockpit. During his tenure he saw the flight engineer removed (despite the protestations of the Union), leaving two bodies to do the same work as 3, and occupy the cockpit.

    I vividly remember him telling us that this was a mistake. He went on to describe how many “egotists” worked as pilots, and it was often the 3rd. man in the cockpit that was used to settle “arguments” that occurred during the flight.

    It’s a shame, really.

  11. How would an enterprising woman be able to tell the difference in a $19,000 per year pilot versus a 150K pilot?
    It seems like you choose the pilot salary to suit the story. (No offense)

  12. @Jim,

    Wouldn’t advancements in navigational ops have been a leading cause in taking out the flight engineer?

  13. Jive Turkey: “How would an enterprising woman be able to tell the difference in a $19,000 per year pilot versus a 150K pilot?”

    She would have to refer to and learn that a captain has four stripes on his or her jacket. Then she would have to look to see if the pilot had an insignia reading “Delta,” “United”, or one of the other major airlines (as opposed to a regional airline such as “Air Wisconsin” or “SkyWest”). A junior captain’s salary is lower than that of a senior captain, but of course the target’s income will grow with the child. Unionized airline pay is pretty easy to find. is one example. Where it says that a 777 captain earns $248/hour in Year 1, multiply by 1000 to get a rough idea of the annual salary (pilots fly close to 1000 hours per year plus get per diem payments that could also be counted toward a child support plaintiff’s entitlement).

    Jurisdiction is important. The most lucrative state in which to target an airline captain is Massachusetts (about $920,000 in revenue when the defendant earns $250k). New York ($892,500 in revenue) and Wisconsin (same rate as New York but for 18 years instead of 21) are also good. So if the initial meeting is in Newark, for example, the future plaintiff can make a lot more money if the pilot can be induced to hop over to Manhattan. Under the Uniform Interstate Family Support Act one of the factors under which a state can exercise its jurisdiction over a nonresident is the nonresident engaged in sexual intercourse in the state and “the child may have been conceived by that act of intercourse,”.

    [Moderator’s Note: the time of this posting was edited for sort order]

  14. It sounds like married men shouldn’t be allowed to become pilots at all. There is an infinite supply of confirmed bachelors in Silicon Valley who could do the job with no marriage induced issues. Not only are there the disastrous consequences of a divorce, but the system is so rigged to guarantee the only people who get married are the mostly likely to divorce & make money for lawyers, all married men have to be a bit psychotic in the 1st place.

  15. Jim,

    Thanks. I have memories of boarding planes as a child and seeing three people in the cockpit (I was a child in the 70s and early 80s, I’m not sure if children are allowed to visit cockpits now). Thanks for confirming that my memories are not false, and yes, removal of the third man was a really bad idea.

  16. Since the flight engineer is not needed to fly the plane anymore, you wouldn’t have to hire a trained pilot to do his job. You could just get a mediator instead. Then again, since you can hire fully trained pilots for $19K per year, they might be cheaper than trained mediators.

  17. I know you love the divorce thing, but it doesn’t seem to be relevant here. The two examples of pilot-induced crashes were foreign, and thus not subject to our draconian divorce laws. Indications seem to suggest that in both cases, there was a direct threat to the Pilot’s career.
    Isn’t the way that the Aviation Medical Community works a little counterproductive. My understanding is that, if one were depressed (or had some kind of concealable heart problem), and one were part of an aircrew, it would open a huge bag of worms to go see a doctor and begin taking medicine. (And the reaction to this incident will likely not help things in this regard) I certainly don’t want a suicidal guy flying me and my loved ones around, but neither does it seem like a good idea to bar anyone that’s ever been depressed from flying. Most depressed people are not suicidal, and most suicidal people do not end up trying to take anyone with them. Is there something we could do with experienced pilots that have an issue with their medical that isn’t career death? (I know you’ve floated the idea of intensive remote-co-piloting, why couldn’t we make that the cushy, well paid gig that pilots emeritus get to go do when their health starts to fail or their marriage is on the rocks?)

  18. Here is another consequence of running budget airlines; you may not be able to afford pilots with all four limbs. This real-life “Monty Python” story concerns the British budget airline Flybe. One journey to Belfast nearly ended in grief when the pilot’s artificial arm came off.

    All lived to tell the tale. To my considerable surprise, BBC Radio 4 News wheeled on an “expert” to reassure us that there was no problem about Mr Three Limbs flying a load of passengers, provided he had passed all the required competence tests. As one wag put it, there was no real problem – he just asked the co-pilot to lend him a hand.
    Admittedly there was the brilliant WW2 fighter ace Douglas Bader, who despite having two artificial legs, flew Hurricanes and Spitfires and was credited with over twenty kills. But, as his biographer pointed out, there was an implicit understanding that he would be allowed to fly only single seater fighters such as those. He would not pilot large warplanes filled with other people’s sons.

  19. SuperMike: The latest news reports suggest that the suicidal/homicidal pilot was in fact about to join the club of aviators paying child support for seldom-seen children. His girlfriend was pregnant and she had dumped him. German child support numbers aren’t high enough to get an American plaintiff excited, nor are they high enough to exceed the returns from going to college and working, but they would result in a substantial spending power reduction for a junior pilot and also presumably crimp his ability to find a new mate.

  20. Letting pilots bring dangerous items within the secured area opens at least one terrorist opportunity : they can give it to an accomplice terrorist passenger, targeting a plane other than his.

    I would guess that recruiting an pilot as a smuggler for terrorists, especially one earning $20K a year, and if it doesn’t required himself to be killed in the process, isn’t impossible.

Comments are closed.