Hacker News Reader

Comment by XorNot | Hacker News Reader

EDIT 3: God damn it - forget it. The rate is per 100,000 - the site I got it from was terrible [1].

So the numbers don't work at all - the death rate should be about ~0.8% in any given year for the full population going by [2] (798.8 deaths per 100,000 in 2022).

Just to elaborate: my thesis before I edited this post was that if you have a large number of whistleblowers, the chance of someone in the population dying of normal causes over a given time frame might become quite large.

[1] https://data.worldbank.org/indicator/SP.DYN.AMRT.MA

[2] https://www.cdc.gov/nchs/nvss/vsrr/mortality-dashboard.htm

View on HN

I agree that it doesn't make much sense to kill someone after they've testified, but that 1.6% figure (I think you inadvertently omitted a decimal point) doesn't account for age. Joshua Dean was 45 (0.4% chance of a 45 year old male dying in a given year). John Barnett was 62 (1.5% chance).

Yeah I was way off on the death figure, but the point is relevant: the risk of a large population and a low probability event is that the low probability event can become a near certainty due to normal causes.

~30 whisteblowers[1] over 3 years is a fair number of people in your chance pool - not guaranteed, but there's confounding factors (i.e. stress, disruption to work and family life etc.)

My point was meant to be that saying "there's a chilling effect" is also creating the chilling effect by implying it's there, when in reality it can just be something more akin to the Birthday Paradox.

[1] https://www.aljazeera.com/economy/2024/4/19/boeing-subject-o...

> ~30 whisteblowers[1] over 3 years is a fair number of people in your chance pool - not guaranteed, but there's confounding factors (i.e. stress, disruption to work and family life etc.)

People who already feel like they have nothing to lose (which I imagine would be correlated with higher chance of death, with causation in either direction) might also be more willing to become whistleblowers in the first place. Doesn't seem a factor in this case though.

The issue is that becoming a Boeing whisteblower is a 1-way gate: once you are one, you're a member of that group for the rest of your life.

Which means if you have a lot of whistleblowers, because you've got a lot of problems, then you have this ever-expanding pool of people who are "Boeing whistleblowers". Which means the probability a Boeing whistleblower dies in any given year goes up as the pool expands.

Which is fine and normal, unless the narrative question we're answering is: "is Boeing killing whistleblowers to deter future whistleblowers?"

And one conclusion is: "no, but they're being deterred anyway, because everyone is insisting they were intentionally killed". Basically the "chilling effect" is the conspiracy theory, not any actual action.

Though I would note a better way to see if they were doing it would be to look at the death rate of current Boeing employees, since if you were killing whistleblowers that's what it would presumably look like - people presently employed by Boeing dying under explainable circumstances without ever taking any action.

> Though I would note a better way to see if they were doing it would be to look at the death rate of current Boeing employees, since if you were killing whistleblowers that's what it would presumably look like - people presently employed by Boeing dying under explainable circumstances without ever taking any action.

Well, that wouldn't have much deterrent effect. I don't think it's likely that Boeing is knocking off whistleblowers, but if they were, it would be more effective as a deterrent for them to do it once the person became well-known.

"That is, in any given recent year, if you take a sample of 1000 random males, within a year, almost 16% of them will have died."

Surely this can't be true.

You'd be correct they're not - it's much lower - per 100,000[1]. Should've cross-checked with birth data.

The underlying point though is valid: whether the rate of people dying is particularly high depends on quite a few factors. The estimate for 2023 in Q2 is 944 / 100,000 = 0.9% or so. So within a population of 100, you'd still be unsurprised to find 1 person has randomly died of some cause. So if you take say, the 30 presumed people who have complained about Boeing over the last 3 years and tracked them as a quorum...

[1] https://www.cdc.gov/nchs/nvss/vsrr/mortality-dashboard.htm

OP's original statement of "The death rate in the United States per 1000 male adults as of 2022 is 163." comes from [1]. That statistic is defined as, "Adult mortality rate, male, is the probability of dying between the ages of 15 and 60--that is, the probability of a 15-year-old male dying before reaching age 60, if subject to age-specific mortality rates of the specified year between those ages." [2]

[1] https://data.worldbank.org/indicator/SP.DYN.AMRT.MA?end=2022...

[2] https://databank.worldbank.org/metadataglossary/health-nutri...

And if you assume that death rate is age agnostic between 15 and 60 (it's not of course, but bear with me), then this allows you to calculate the yearly chance of death as a US male adult as 1 - ((1000 - 163) / 1000)^(1/(60 - 15 + 1)). This comes to 0.386%, or ~1 in 259.

> The death rate in the United States per 1000 male adults as of 2022 is 163.

Where did you get that number? https://en.wikipedia.org/wiki/Demographics_of_the_United_Sta... and https://www.cdc.gov/nchs/fastats/deaths.htm are an order of magnitude lower for the general population.

Yeah I misread the graph - I've corrected the post, although the errors sort of cancelled towards the overall result anyway. With a CDC reported rate of ~0.8-1% per year, then in a group of 10 you'd expect 0.1 deaths if they were suitably random. In a group of 100 you'd expect at least 1.

They're not random, but it's also a group with one-way membership. John Barnett was a whistleblower from 2017 till his death in 2024. So for 7 years he was a Boeing whistleblower, but every year he is he's still in that 0.8-1% chance of all-cause mortality across the general population (likely more given specific non-murder related risk factors).

EDIT: I guess the reason I'm posting up a storm on this is not to defend Boeing, but that there's an actual, identifiable harm to the conspiracy theory narrative which is that it's all fun and games unless you work at Boeing and see evidence of wrongdoing. We the public want and need you to come forward, and a storm of commentary which says "lol Boeing just straight kills people" is all good hot-take fun till you're in an actual position where it might feel quite real. Even if the odds were much worse, going on the internet and saying "this was definitely murder to create a chilling effect* is helping create a chilling effect.

Surely the average age of the whistleblowers is quite a bit lower than the average male of the whole male pop, or did you take that into account?

> Plug that into the fact that it doesn't work to kill someone after they have testified

This way deters future whistleblowers.

If they got rid of whistleblowers before they blew the whistle, no one would hear about it. It would not be widely reported. Just another random employee who died.

Wouldn’t they want to carry out their whistleblower deterrence operation before there were already 30 other whistleblowers?

Get rid of whistleblowers before they blow the whistle? So they just start killing random employees who look suspicious?

No. If they were trying to accomplish what you were suggesting, you kill the first whistleblower early on to prevent others. That motive doesn’t make sense after there are already 30 whistleblowers.

Their unpredictability makes it more terrifying to future whistleblowers.

Reply on news.ycombinator.com