the trolley problem meets climate change.

[madd0ct0r]

If I can demonstrate that 1 million people's cumulative actions in country A will add 0.002% to the effect of global warming and over the next decade result in 2 billion (thousand million) deaths (or equivalent), is it morally justifiable to kill them?

[Borgholio]

Logically, yes. Morally, no.

In other words, mathematically it makes sense, but ethically...murdering hundreds of innocent men, women and children cannot be justifiable in any way. A better option would be that if you have undeniable proof that 2 billion people WILL die...use that proof to convince that original million people (or more) to change their lifestyle to not have as big of an impact.

[Formless]

If you had the power to just kill them, is it not also likely that you have the power to enact draconian laws, and enforce them? Or simply make yourself some kind of authoritarian ruler, and then cut their carbon emissions through your rule? Or dismantle/destroy high emission coal firing electrical plants, which are by far the single biggest way we contribute to the rise of greenhouse gases in the atmosphere, and replace them with something more Eco-friendly like nuclear power?

If you have the kind of power to enforce such drastic actions, there are many lesser drastic actions you can still take before mass murder becomes necessary. Its not the "trolly problem", because like most real life situations, there is no artificial "thought experiment" circumstances stopping you from finding a third option or a middle ground between your two choices.

[Simon_Jester]

Another reason not to make this kind of calculation is that it's unreliable, especially in the limiting case where the probabilities on one (or both) sides are very large.

A more realistic example where someone actually DOES make this calculation would be, say...

"If we don't screen for X Syndrome, 1 in ten million people will die of it. If we do screen for X Syndrome, 1 in 10000 people will show up as a false positive for the syndrome. Treatment of the syndrome requires surgery that would have a 0.05% risk of killing the patient. Therefore, out of a population of 100 million people, we can either (statistically) kill 10 people by not screening for the disease, or (statistically) kill 5 people by screening for the disease."

Here, the policy decision is at least broadly credible and in the practice of medicine does involve making such choices, although usually it's done patient-by-patient and not for a whole population.

Now, if we make one of these choices on the basis of utilitarianism, we're going to err on the side of screening: five less deaths is good, right?

But notice something- if your calculations are incorrect on either side, you may have committed a serious utilitarian evil.

Maybe X Syndrome is less common than we thought. Maybe X Syndrome is about to become less deadly due to the invention of a new medication, so that fewer people carrying the syndrome die of it... making a high-risk early treatment less important.

Maybe the effect of the false positive test results on people who do NOT die of the surgery are worse than we thought. Maybe a few of the people who incorrectly 'find out' they have X Syndrome become depressed and commit suicide.

Maybe the cost of all this screening absorbs medical resources that could have been used to save other lives in other ways.

It's hard to know, and hard to predict.

So if we're prudent, we will avoid making such calculations about marginal cases. Risk assessment isn't a very exact science, and it is very bad at predicting the future based on past performance. So making a decision to kill people purely because of a statistical conclusion is dangerous, because of the risk that the conclusion will turn out to be wrong.

How would you feel to be the man who killed a million people because of a 0.2% probability... only to learn that it was actually a 0.02% probability and your statistician misplaced a decimal point?

__________

DISCLAIMER:

This is a contrived example but is at least broadly realistic, because it comes up- not testing for a medical condition has consequences, but tests have false positives and the consequences of a false positive are themselves sometimes life-changing or even fatal.

In real life, we'd be scrambling for ways to identify at-risk populations and screen only those people, which reduces the risk of false positives and people dying in surgery. Or for ways to make the surgery safer, same idea. As Formless notes, in real life there's a middle ground.

[Terralthra]

Simon, a great real-life example to replace the hypothetical one is the issue of treating strep throat with penicillin/amoxicillin. People present with pharyngitis and are tested for strep. When a positive result is obtained, we typically prescribe a course of antibiotics. These are unnecessary to treat the strep throat, as it will typically resolve on its own, only taking a day or two longer than the antibiotics cure it anyway. We give the antibiotics more to prevent the rare complication rheumatic fever, which is potentially deadly.

However, somewhere around 1% of people are allergic to penicillin, with 0.01% having a life-threatening anaphylaxis. No more than 3% of untreated strep cases result in rheumatic fever.

On top of that, some people (up to 25%) are functionally immune to strep throat, and have it as part of their throat flora at all times, and thus present with pharyngitis, test positive for strep and are given antibiotics, but actually have viral pharyngitis and do not benefit from antibiotics at all.

So, from a utilitarian perspective, what do you do with patients who present with pharyngitis and test positive for strep? Antibiotics?

This has every bit of the complex utilitarianism of the scenario you present, including that the "correct" answer changes based on small percentage point differences in the risks of allergy, rheumatic fever, and strep immunity.

[K. A. Pital]

Logically, yes. Morally, no.

In other words, mathematically it makes sense, but ethically...murdering hundreds of innocent men, women and children cannot be justifiable in any way. A better option would be that if you have undeniable proof that 2 billion people WILL die...use that proof to convince that original million people (or more) to change their lifestyle to not have as big of an impact.

What if the 1 million are pompous entitled assholes who would just shrug and say 'so what if the dog lay on the hay, doesn't mean it had any right to it in the first place'?

[Borgholio]

Logically, yes. Morally, no.

In other words, mathematically it makes sense, but ethically...murdering hundreds of innocent men, women and children cannot be justifiable in any way. A better option would be that if you have undeniable proof that 2 billion people WILL die...use that proof to convince that original million people (or more) to change their lifestyle to not have as big of an impact.

What if the 1 million are pompous entitled assholes who would just shrug and say 'so what if the dog lay on the hay, doesn't mean it had any right to it in the first place'?

Well if those 1 million people were all grown adults who have chosen to be the biggest, greediest, most selfish assholes on the planet and deliberately ruin life for everybody else...I still wouldn't want to commit mass murder. But relocation to a large, controlled reservation in the middle of Nebraska where they couldn't have enough of an impact to kill those 2 billion people...that would be an acceptable option for me.

[K. A. Pital]

Well if those 1 million people were all grown adults who have chosen to be the biggest, greediest, most selfish assholes on the planet and deliberately ruin life for everybody else...I still wouldn't want to commit mass murder. But relocation to a large, controlled reservation in the middle of Nebraska where they couldn't have enough of an impact to kill those 2 billion people...that would be an acceptable option for me.

Me neither, but I would certainly be very concerned (being able to cause that much global warming in such small numbers would mean a lot of power and clout and industrial equipment). It is unlikely you would have the ability to stop them unless you attacked them and destroyed the power structure that allows them to do what they do by force. It doesn't follow that they all have to be killed, of course, just as it didn't follow that all Germans had to be killed cause of Hitler. However, some casualties during the demolition of the hierarchy or power system that allows them to do it are inevitable.

[Metahive]

Another question, is a person who can reduce a question about life and death of this magnitude to a simplistic mathematical formula actually qualified to make such decisions?

[K. A. Pital]

Another question, is a person who can reduce a question about life and death of this magnitude to a simplistic mathematical formula actually qualified to make such decisions?

No, not really, but aren't we making these scenarios to evaluate the possible outcomes? Utilitarianism has its flaws. One of them is the ability to mathematically justify slaughter in case of scientifically solid predictions of binary outcomes. Some other system would also have problems, namely, it would disallow direct murder of the determined culprits but at the same time be fine with indirect murder of a huge fraction of human population. This has been the case for a long time when economic pressure was applied to huge populations and people did nothing to stop it due to the indirect nature of the process. The scenario is really very much similar to the total war incidents that occured in real history - was the bombing of Axis countries justified? Millions of Axis civilians died - but tens of millions of civilians died in the nations that were attacked by the Axis.

In reality such decisions are usually made by people even less prone to pondering. Like Buck Turgidson, that level of humanity.

[Metahive]

But the question is, can these simplified scenarios actually deliver a solution befitting the much more complex RL scenarios? Take one of the simplest forms of these, the people are starving and food relief is so far away that all will have starved until then. Solution, feed one half of the people to the other half, so the people will be fed and there's also less people to feed in total. How applicable is this to a real life starvation scenario?

[Terralthra]

But the question is, can these simplified scenarios actually deliver a solution befitting the much more complex RL scenarios? Take one of the simplest forms of these, the people are starving and food relief is so far away that all will have starved until then. Solution, feed one half of the people to the other half, so the people will be fed and there's also less people to feed in total. How applicable is this to a real life starvation scenario?

It's...exactly what people do in that situation?

[Metahive]

I'm talking about this as something enforced by authorities. During the Great Leap Forward starvation was such a large issue that people supposedly swapped their children with other families to kill and consume them, but it wasn't something the party told the people to do (since they were in denial about it anyway). The question is, can such scenarios, or even the limited RL examples you gave be used to justify an official "cannibalism directive" of some kind?
This is the sort of applicability I'm questioning.

As for the original scenario in the OP, is wouldn't there be an obvious overlap between the two populations? GW isn't exactly picky on what areas it affects.

[K. A. Pital]

The government does not always follow strict utilitarianism (see my comment about them not being ubercalculating machines but rather people, often too simple people).

[madd0ct0r]

As for the original scenario in the OP, is wouldn't there be an obvious overlap between the two populations? GW isn't exactly picky on what areas it affects.

The most vulbnerable people are poor, living in the tropics on coastlines and estuary flats. They tend not to be huge consumers of fossil energy

[Alyrium Denryle]

I'm talking about this as something enforced by authorities. During the Great Leap Forward starvation was such a large issue that people supposedly swapped their children with other families to kill and consume them, but it wasn't something the party told the people to do (since they were in denial about it anyway). The question is, can such scenarios, or even the limited RL examples you gave be used to justify an official "cannibalism directive" of some kind?
This is the sort of applicability I'm questioning.

As for the original scenario in the OP, is wouldn't there be an obvious overlap between the two populations? GW isn't exactly picky on what areas it affects.

There is a difference between being "good" and being "necessary". Ethical systems are tools. Different ethical systems work better under different conditions. Various forms of deontology work very well for interactions between individuals and when governments have to concern themselves with individuals. It fails very badly when effects are indirect, when Lesser or Two Evils conditions apply, or when problems exist on a massive scale.

Utilitarianism is the exact opposite. Very good for megapolicy, bad interpersonally.

But the thing with Utilitarianism is that it deals very well with Lesser of Two Evils scenarios. It cares about a ratio, and the lesser of two evils is still the better ratio, even if none operate under the delusion that the outcome is good. It wont paralyze you into potentially catastrophic non-action like deontology will.

..................

As an aside to everyone. Stop being fucking squeamish. There is a reason ethical dilemmas are presented, and it is not to give you an excuse to try to wiggle your way out of the scenario. Rather, to force you to confront a difficult question. So confront it, rather than attempting to dodge it.

The answer is Yes. You kill those million people. Provided that another million will not compensate for their cumulative action.

[Simon_Jester]

I should think the first-stage answer is "check very fucking carefully with your statisticians." And the second-stage answer is "check very fucking carefully to see if those million people have armed friends who might completely derail and destroy all your plans out of sheer vengefulness."

[Alyrium Denryle]

I should think the first-stage answer is "check very fucking carefully with your statisticians." And the second-stage answer is "check very fucking carefully to see if those million people have armed friends who might completely derail and destroy all your plans out of sheer vengefulness."

In the real world yes, but these scenarios inherently assume you are somewhat Q Like. Either that, or they set up the scenario properly.

Here, lets construct an extensionally equivalent scenario to see if it is more palatable.

There is a man standing at the entrance to a (non-US) football stadium. He has a detonator in his hand that will activate an explosive device that will kill 1000 people. Your mind has been transported by way of telepathy into the body of a police sniper who has apparently developed narcolepsy at a very inconvenient time. . You have an appropriate rifle trained on the man's brainstem, it is zeroed in and there is no crosswind. Any round you fire WILL hit his brainstem, and there will be no adverse consequences to the sniper, because as far as anyone else is concerned he was doing his job. He is less than a second away from pressing The Button, you have no opportunity to re-aim.

Do you kill him?

How is this scenario any different ethically from the scenario in the OP?

Now for a slightly different scenario.

Same football stadium. Similar madman. Save that this one has gone through a great deal of effort to make his killing of 1000 people indirect. He has put in place a concrete blockade that will extend from the sidewalk right in front of one of the doors in five seconds unless his heart stops, which freezes the timer. He has also put a small incendiary device in place set to explode in ten seconds. It wont kill anyone, but it will cause a panicked evacuation. The obstruction to that one door will kill 1000 people or thereabouts due to crushing and trampling. If the obstruction is not in place, no one dies because the stadium was designed to facilitate efficient and safe evacuations. Assume you know this due to infallible foresight. You have been transported by act of Q into position with an identical sniper rifle.

Do you kill him?

How is this any ethically different from the OP scenario?

[Darmalus]

Because in both cases the madman is intentionally trying to kill people, directly or indirectly.

Who, exactly, is intentionally trying to raise the global temperature?

[Alyrium Denryle]

Because in both cases the madman is intentionally trying to kill people, directly or indirectly.

Who, exactly, is intentionally trying to raise the global temperature?

The executives of oil companies to name a few. Or rather, they are committing actions which THEY KNOW will increase global temperatures, and doing so for private gain. To the point that they have engaged in elaborate deception campaigns to convince the public that the danger does not exist.

Once people start to die, they could (as in, maybe) actually be charged with 2nd degree murder in states where Depraved Indifference to Human Life applies as a subset. Voluntary Manslaughter or perhaps Negligent in most other US jurisdictions. Of course, I may be biased in my read of the law.

There is precedent for it. Men being convicted of attempted murder for knowingly slutting about while HIV positive and not informing partners (in some cases knowingly lying about their HIV status) and not using protection.

[Darmalus]

The executives of oil companies to name a few. Or rather, they are committing actions which THEY KNOW will increase global temperatures, and doing so for private gain. To the point that they have engaged in elaborate deception campaigns to convince the public that the danger does not exist.

Once people start to die, they could (as in, maybe) actually be charged with 2nd degree murder in states where Depraved Indifference to Human Life applies as a subset. Voluntary Manslaughter or perhaps Negligent in most other US jurisdictions. Of course, I may be biased in my read of the law.

About as likely as me wining the lottery and never having to work another day in my life. Possibly less likely. Huge chains of leadership selling perfectly legal substances that an economy depends on for generations. Economies that will STILL depend on those substances, even when LA and NY are under water, simply because the lag time between deciding to do something and the solution being put into place. There won't be any dramatic event we can point to, just deserts growing and water levels rising at rates to slow for humans to really notice day to day.

Responsibility is very diffuse. If anyone gets punished, their picture should be in the dictionary next to "scape goat" since there will be a few billion hands with blood on them.

If those hypothetical million people were killed, I'd be surprised if there was a single oil executive among them. It would probably mostly consist of politicians, city planners, accountants, marketing people, advertisers, farmers and random vocal citizens. The people you never notice, but eventually contibute to the logical but shortsigted decisions to make cities more spread out, farmers who sell their land to become suburbs, the marketer who hits on a really good way to sell SUVs, and so on.

There is precedent for it. Men being convicted of attempted murder for knowingly slutting about while HIV positive and not informing partners (in some cases knowingly lying about their HIV status) and not using protection.

I don't know what laws it would fall under, but I'm pretty sure that intentionally spreading a deadly disease is illegal. Marketing and selling perfectly legal oil is not a crime.

[Simon_Jester]

The key to Alyrium's argument is that if you're poisoning or hurting someone on purpose, it's wrong whether it's illegal or not.

Keeping track of the difference between 'legal' and 'ethical' is a big thing in ethics debates...

[Darmalus]

The key to Alyrium's argument is that if you're poisoning or hurting someone on purpose, it's wrong whether it's illegal or not.

Keeping track of the difference between 'legal' and 'ethical' is a big thing in ethics debates...

One of the reasons I'm less than impressed with internet ethics is it always seems to boil down to condemning people who made decisions decades ago for not having godlike omniscience and foresight. When you start talking mega-deaths to save potential future people that may or may not exist, I'm pretty sure you're the monster.

The original trolley problem works because it is an immediate problem on the human scale. This scenario sounds like some just wants to get some revenge, not make the right decision even is millions for billions works out mathematically.

[Simon_Jester]

I'm actually agreeable to this- I don't have enough faith in the power of prediction to be that confident of it. Not confident enough to do a bunch of math, go "therefore killing these one million people will save 1.5 million people elsewhere," and pull those one million triggers myself.

It's one thing when there's a massive, orders-of-magnitude difference between the two evils you're comparing. Or when both evils are present-day things whose scope can be calculated accurately. Or when there's some overriding principle involved we can adhere to a la rule utilitarianism.

But by itself... I'm not enthusiastic.

[Borgholio]

The only way this would work really is if it wasn't as abstract (can't think of a better word) as climate change. Say those 1 million people were a breakaway population of scientists, biologists and ex-military that took over a Pacific island and are developing a bio-weapon that can kill billions. In a case like that, nuking the island flat would probably be a no-brainer.