59 replies
12h38m
Narrator: A new car built by my company leaves somewhere traveling at 60 mph. The rear differential locks up. The car crashes and burns with everyone trapped inside. Now, should we initiate a recall? Take the number of vehicles in the field, A, multiply by the probable rate of failure, B, multiply by the average out-of-court settlement, C. A times B times C equals X. If X is less than the cost of a recall, we don't do one.
Business woman on plane: Are there a lot of these kinds of accidents?
Narrator: You wouldn't believe.
Business woman on plane: Which car company do you work for?
Narrator: A major one.
While I think I get your point, I also believe that the background is very important. If your company applied standard engineering procedures, that's critically different from building a new product and actively hiding the differences with respect to the last one.
It is a stretch to call the MAXes a new product, even though it turned out to require more extensive work than envisioned, but the bit about actively hiding the differences is right on point.
There’s a bigger difference between the MAX family and other versions of the 737 than there is between all the different crossovers a car company pumps out.
Boeing 737 MAX 7 is 138,699 lbs vs 737 800 @ 90,710lb vs 737 100 @ 61.994 lbs. Hell the new 737 MAX 10 is ~203,000 lb.
This is not the issue that is raising questions of criminal liability.
It’s definitely part of the issue.
Boeing was trying to pretend scaling the aircraft again and again wasn’t significant due to regulations and physics eventually disagreed.
Regardless of how one characterizes the magnitude of the changes, if the unanticipated difficulties had been handled openly, there would be no question of criminal liability. Furthermore (though it is also beside the point), physics has not ruled out the MAXes.
In addition, the door-plug issue is tied in here on account of the 2021 deferred prosecution agreement, which also followed from Boeing's duplicity over the MAX issue.
You’re skipping over the first half of my statement. If the unexpected difficulties had been handled openly they would have needed to go through more regulatory hurdles. Physics didn’t put them into some kind of catch 22 situation the aircraft could have been safe, it just couldn’t be safe while playing games with regulators. That’s where criminal liability shows up.
As far as I can tell, I'm just skipping over the issues which may be true, but are beside the point, but then, it is not clear to me which statement's first half you think I am skipping over.
The point I have been making all along is that, regardless of what led Boeing to the point of choosing to hide or misrepresent the situation, it is the choice to do so that turns this into a potentially criminal matter. With the same physical/technical problems but proper and timely disclosure of the issues during development, by far the most likely outcome would have been a delayed program delivering MAXes substantially similar (in both construction and operation) to the ones which are certified and flying today. After that, even in the unlikely event that the crashes had occurred, criminal prosecution would be unlikely, and certainly not on the basis of the facts for which it is currently being considered.
Except they aren’t beside the point.
The physical aircraft would have been similar, but airlines would have spent 10’s of millions more on training which makes a real difference to them and thus sales.
Boeing could have released the aircraft on exactly the same date while complying with the spirit of relevant relations though at higher costs, but the product would have been meaningfully worse from a sales perspective. Even today regulators have allowed Boeing and the airlines to treat the 737 MAX family as much more closely related to earlier 737’s than they actually are.
All of these things are beside the point here because none of them form the basis on which criminal proceedings are being considered by the DOJ. Without the information hiding and misrepresentation, there is no basis, regardless of either the physics or the economics of the issue.
In addition, even if it is true that regulators have now allowed Boeing and the airlines to treat the 737 MAX family as much more closely related to earlier 737’s than they actually are, this is not the basis of the DOJ's investigation, either.
Criminal proceedings in the US care about motives. It would be vastly harder to bring a case like this if there weren’t incentives to ack as they did.
Regulators aren’t at issue, but trying to avoid regulatory scrutiny is. Or as is often said it’s the coverup that they get you for.
The motives were there regardless of how Boeing chose to act. It's the chosen act that is potentially criminal. The DOJ is not in any doubt as to whether or not Boeing had a motive, and it is not considering at all the question of whether regulators have now allowed Boeing and the airlines to treat the 737 MAX family as much more closely related to earlier 737’s than they actually are, regardless of any parties' motives in that regard.
A and B is false if B is false, but that doesn’t make A irrelevant.
Is the size the biggest difference or the fact that the center of gravity is totally different? So much so, they created an "hidden" software program to counter the CoG difference so the pilots feel like it is in the same place they are used to?
No, CoG has approximately nothing to do with it.
The problem with the Max is aerodynamics: Civil airplanes are designed with negative aerodynamic feedback, so the nose-down torque (“pitching moment”) increases as the AoA increases. But on the Max at very high AoA this feedback torque becomes somewhat smaller due to interaction of airflow around the engines and the wing. That by itself is not a problem because the nose-down torque still exists and airframe is still stable.
There is a regulatory requirement that the amount of pilot’s control force required to maintain a given AoA must be a non-decreasing function of AoA[1]. Due to the Max’s aerodynamics, this requirement is not met.
Boeing’s MCAS was a bandaid to make the plane meet regulations by applying nose-down trim while at high AoA. The trim results in a higher yoke force, so the plane meets the requirement. A better method would have been a “stick pusher” (which have been used as stall prevention devices for over 50 years, though not on the 737) or addition ventral fins (like on the Beech 1900). But either of those would have probably required recertification
[1] The purpose of this is to reduce pilot-induced oscillations: see what happened with AA flight 587, where the lack of a similar requirement for rudder pedals and yaw led to PIO which ultimately resulted in the tail falling off an Airbus.
No, you are confusing the empty weight of the 737-800 with the mass gross takeoff weight of the 737 max 7.
The correct comparison of MGTW is 174,200 (-800) vs 177,000 (max 7).
Similarly, the -100’s MGTW was 110,000, not 61,994. Also, the only source I can find for the max 10 is 197,900, not 203,000.
Edit: so the plane grew by 80%, not 320%
Like Toyota's car with a “simulated stick shift”? https://www.caranddriver.com/news/a45754176/toyota-manual-ev...
Sure, simulating the stall characteristics of an aeroplane is several orders of magnitude worse than simulating the stall characteristics of a car, but… still.
It doesn't sound like Toyota is trying to hide anything, they are preparing to sell something they think people want.
(The feature sounds stupid to me)
I've been eyeing the Ioniq 5 N which has the paddle shifting rev-limiting simulation stuff as well. Most of the reviewers really enjoy those features on a track. When coming up to a turn, the simulated RPM gives you much more contextual information than a pure electric experience does. The main point that kept coming up is the RPM levels when approaching a corner. They could tell if their speed was right in the approach based on the sound the engine was faking which is how they would drive a manual ICE car through the same track. Of course you can turn all that off and soften the suspension right back up in order to get the comfortable daily driving most folks are after. I don't think it belongs on most EV, but it certainly seems to enhance some. At least while we're still more familiar with the trappings of ICE performance vehicles.
I would guess that it will vanish when someone comes up with a similar feedback mechanism that is better aligned with the performance characteristics of the powertrain.
I think the technology will explode when you can download a tune from classic or cool cars and have the sound and performance map of a Ferrari or Aston Martin when you go out for a fun drive on weekends.
It’s not about max speed but enjoying the experience of driving.
99.9% of cars with cvts pretend to switch gear instead of just acting like a CVT.
This might seem calculating and cold, but that's basically how government agencies work. It doesn't make sense to spend unlimited amounts of money to save a life, so government agencies have some sort of a dollar value on a life[1], above which where they won't bother doing interventions. For instance if it takes $15M to save a life, and that's above the DoT's estimate of $12.5M, then they won't bother. That's not to say that Boeing is in the right here, but if they were in the wrong it's not because "If X is less than the cost of a recall, we don't do one", it's because they grossly undervalued a life in their calculations.
[1] https://en.wikipedia.org/wiki/Value_of_life#United_States
Government agencies use the statistical value of a life to weigh the benefits of a regulation against the costs to the public - the same people who would benefit from the regulation. Would a new safety requirement for cars be worth it if it added $500 to the price of every new car? They might use the value of a life to help answer that question.
This is very different than the recall story, because in that scenario there is no cost to the public at all, only to the corporation. When government agencies make rules about product recalls, they do not try to balance the benefit to the public against the cost to the recalling company, that would be insane.
See the DoT's own guidance here: https://www.transportation.gov/sites/dot.gov/files/2021-03/D.... They do not at all use this figure in the same way that a profit-driven entity does.
The word "individuals" is very important in this definition.
...assuming they don't pass it on to the customers
That would be illegal, and "the public" is much larger than just the customers of a particular company. This is beside the point though. The government asks "would the average person be willing to pay $x to lower their chance of death by y%". The corporate executive asks "would I be willing to pay >$0 to lower their chance of death by y%". "Some of you may die, but it's a sacrifice I am willing to make."
As the other commenter has mentioned, it's passed on via higher prices in the future.
Okay, but surely you don't agree that Boeing should spend infinite amounts of money making their planes safe? For instance we don't install backup engines on the off chance that all 2 engines fail. That's all I'm trying to argue, that the cold calculation/cost benefit analysis as mentioned in the OP isn't where Boeing went wrong, it's that they they undervalued the value of a human life. This was specifically mentioned in my original comment.
What I am trying to argue is that valuing a human life accurately doesn't matter if you're weighing it against your own profits. The correct comparison is against societal benefit, profit is completely morally irrelevant.
I agree that the ceiling for how much money you could spend trying to make an airplane perfectly safe is infinite, so by definition they have to stop somewhere. However, I disagree that finding that line of where to stop has anything to do with the statistical value of a human life.
For example, imagine the Boeing CEO says "we could spend $20 billion on R&D and manufacturing of a new safety system that would on average prevent 1 crash per year, but a 737 MAX carries 200 people, at $15M per life that's only $3 billion in value, so it's obviously better for us to skip it and pass that $20 billion on to shareholders". This would be criminal, and not because their value of a life is off and they got the math a little bit wrong. It's criminal because they are consciously choosing to kill people unnecessarily.
If it is physically and financially possible to make your product safer, you do it, without any thought to how much a life is worth. If it is not possible, because you can't figure out how to solve a problem or it would be so expensive to fix that your business couldn't survive, then you sit down and think about whether it's worth selling your product at all. Are there safer alternatives available? Could a better-funded company fix the flaws you've found? If you determine that your product is important, there are no alternatives out there, and it cannot be made safer, only then do you start weighing the benefits to society against the deaths you expect to occur. I don't really think this is a financial decision involving the value of a life - if you expect your product to kill people then the expected benefits need to be so overwhelming that doing the financial math is unnecessary.
It is not criminal and I have no idea how anyone can think it is.
The crime is in the first paragraphs of the article:
So to recap, they were agreed to implement internal rules that match US laws after killing a bunch of people to avoid criminal liability, _which they failed to do_.
What do you mean it would be illegal? Obviously the corporation prices in all the expected future recalls into their prices. Yes they won't charge you for the recall, but the future costumers will pay higher prices if in the corporation's calculation the chance of recalls (or the cost of doing them) is increased.
Why oh why do I have to read this on HN again and again? The price of a good does not primarily depend on the cost to make it. So no Boeing can't easily raise their future prices to account for the cost of saving lifes, because there are competitors.
If they could have charged more for the car they already would have. A recall only affecting one make or model won't have a large effect on the profit-maximizing price of the car because the value proposition of the car hasn't changed (or got worse because of perceived unreliability). The cost would then be born by the company and not the consumer.
They could make products or services more expensive but then manufacturers with many recalls would end up being prohibitively expensive.
Each person can just choose individually a different manufacturer from the many options, one being "none at all". But you only get one government, it takes a majority to choose, and once chosen you're bound to its directions. So not the same thing at all.
We were taught as part of our CS curriculum that "often, the refusal to assign a dollar value to a human life results in it being terribly undervalued."
Put some CEOs in jail and you won't need to do all those multiplications.
There is no such reluctance to do this in law. I think in philosophy classes, professors feel shy about acknowledging the humdrum reality of insurance, the tax code, and our regulatory bodies continuously assigning dollar values to human lives. It feels good and makes you look like a better person to say things like "the value of a single human life cannot be expressed in money terms." Perhaps this is the case, and what the companies and the governmental bodies are really valuing is the economically meaningful activity associated with a person and not the totality of their person.
Babies are worth a lot of money even at a low imputed earning potential just because they have so much life to live.
There’s always a risk calculus that seems cold to an outside observer.
The issue with Boeing and the MAX is that they sort of used the designation of the plane to for lack of a better term, avoid some of the risk calculations.
With the early issues, a pretty cut and dry type training program would have likely prevented catastrophic incidents. They shaved pennys and set many dollars (and hundreds of people) on fire.
If you run a company where low probability, high impact risks drive the operations of your products, setting trust on fire is going to have a real impact on your bottom line. You go from a trusted, admired company to target of memes about assassinated whistleblowers.
My very amateur understanding is that the MAX would have been dead on arrival if it required recertification. It was a marginal bump in fuel efficiency that still lagged behind Aerobus, so their only real marketing strategy was not having to recertify.
Basically either it didn’t require recertification or it wasn’t worth making. I suspect somewhere along the line they realized that wasn’t practical and they either shipped a shoddy project and maybe burned a pile of money or they gave up and definitely burned a smaller but still sizable pile of money.
Probably also some worry that if they didn’t have a competitive smaller plane, they’d lose market share and there’d be even less interest in their next plane.
Government agencies in the form of rescue/emergency response organisations frequently spend huge sums of money to try to rescue indidivuals.
The US Coast Guard explicitly stated that they "do not associate cost with saving a life" following that fateful Billionaires game of chicken with extreme water pressure.
The problem with this type of thinking is that it often assumes externalized costs, and even at that, what price do you put on “I don’t want to use car of brand X because the car maker is doing sus things and people speaking out about it are dying.” ?
Take for instance the recent train derailments. Are the costs of cleanup n>0 if the company can get their lawyers to successfully argue they don’t need to pay it? For the company it’s n=0 but for society it is n>0.
Doing the maths is simpler when you can disregard many factors and leave only the ones that affect your bottom line.
No it's because they grossly underestimated the importance of brand trust and reputation
In practice, there's a difference between a hypothetical probabilistic death rate, and real people / news stories.
This is from the book Fight Club https://www.goodreads.com/quotes/691547-wherever-i-m-going-i...
More accurately, this is a quote from the Fight Club movie
Honestly this quote is one of Chuck Palahniuk's best dialogues. Recommend checking out his other work like 'Rant.'
Is this referencing a real event or interview?
It's referencing Fight Club
Thank you.
This formula is for the statistically stupid. There is no such thing as average out-of-court settlement. One jury may find your company’s behavior as egregious and decide to drive you out of business by awarding a trillion dollar judgment. Next your business insurance drops you for being uninsurable.
1. There are long tail distributions which have a statistical mean
2. Boeign exists
Reminds me of a business consultant who recently got roasted on TikTok recently because he described his $500k job as deciding whether or not companies should recall products or just weather the class action and liability suits.
Source?
oh so this is what google facebook elon musk microsoft does to collect/share user data then get slammed with affordable fines later. I surprised this quote hasnt come to mind at reading those article
It’s comical nobody gets this reference, sir.
https://en.wikipedia.org/wiki/Ford_Pinto#Fuel_system_fires,_...
What's the alternative? Never ship unless the risk is 0.00? (And yes I get the book reference)
For airlines and many other things you also have to consider the case that people might switch to even worse alternatives if the cost is too high. If airlines have to raise the prices too high for added safety features, people will switch to cars which are several orders of magnitude worse in safety. So airplanes might become safer but society may be overall worse off.
Just C is set to a high enough value, including personal penalties and imprisonment for any executives who failed to stop this decision (including the failure to setup reporting to hear about such decisions).
This 5ypical MBA formula ignores the most important variable.
Z = brand trust