A trolley problem

by John Quiggin on July 26, 2017

I’ve generally been dubious about trolley problems and similar thought experiments in ethics. However, it’s just occurred to me that an idea I’ve tried to express in the economistic terms of opportunity cost, without convincing anybody, might be more persuasive as a trolley problem. So, let’s start with the standard problem where the train is about to kill ten people, but can be diverted onto a side track where it will kill only one.

In my version, however, there is a second train, loaded with vital medical supplies, which is about to crash. The loss of the supplies will lead to hundreds of deaths. You can prevent the crash, and save the supplies, by diverting the train to an alternative route (not killing anybody), but you don’t have time to deal with both trains. Do you divert the first train, the second train, or neither?

Hopefully, most respondents will choose the second train.

Now suppose that the first train has been hijacked by an evil gangster and his henchmen, who will be killed if you divert it, but will otherwise get away with the crime. As well as the gangsters, the single innocent person will die, but the ten people the gangster was going to kill will live.

The impending crash of the second train isn’t caused by anybody in particular. The region it serves is poor and no one paid for track maintenance. If the train doesn’t get through, hundreds of sick people will die, as sick poor people always have, and nobody much will notice.

Does that change your decision?

As I hope at least some readers will have realised, this version of the trolley problem is a metaphor for humanitarian military intervention. The moral intuition supporting such intervention is the same one that would lead to choose stopping the gangster over saving lives of people who would otherwise die as a result of poverty and disease.

As I’ll argue at length if needed, the numbers in the example are stacked in favor of humanitarian intervention. Many such interventions kill more people than they save. Even where they are successful in their own terms, the cost is massively more than that of civilian aid, for a fraction of the benefit.

One final point is that, in reality, the ‘henchmen’ are often conscripted, by force or economic necessity, from the same population as the people whose lives are supposed to be saved by intervention. On any reasonable account, their deaths ought to be weighed in the ledger against any lives saved.



Phil 07.26.17 at 9:25 am

What you’ve omitted to mention is that the gangsters aboard train 1 have with them a stock of deadly plague-spreading biological agents, with which they intend to infect every city and town that their train passes through, causing untold suffering and death. And given that the line train 2 is on was in a bad state anyway – it was last year and it probably still will be next year – it makes more sense for the international c -ahem- the hypothetical pointsman to concentrate on stopping the new and unknown menace of train 1.

(I don’t endorse this argument, but I think yours is vulnerable to it. Simply saying “No aggressive wars! No imperialist interventions! No, not even this one!” and nipping the whole thing in the bud seems more effective to me.)


John Quiggin 07.26.17 at 9:59 am

@Phil “No aggressive wars! No imperialist interventions! No, not even this one!” isn’t any use, since no one is an aggressive imperialist in their own eyes.

If you’re going this way, you need to say simply “No wars! No interventions! No, not even this one!” .

That may be more useful than “Nearly all wars have bad consequences, and there’s no reason to think this one will be different”, but I don’t hold out much hope either way,


afeman 07.26.17 at 11:45 am

On Twitter somebody graphically posed the trolley problem with McCain’s doctor at the switch.


Anders 07.26.17 at 12:05 pm

John, why would a government need to choose between alleviating poverty and disease, and war?

Whilst economically governments ought rationally to think of a military intervention in terms of opportunity cost, I don’t think they ever do.

AFAIK, no government has ever thought of war as subject to resource constraints. There’s always a magic money tree when it comes to invading another country. So it’s not a choice of A or B, surely.


Andre Mayer 07.26.17 at 12:35 pm

It’s always seemed to me that the trouble with trolley problems is that they tend to get into hypotheticals that you can’t really know/weigh properly — here, for example, the ten people the gangsters are supposedly going to kill. As a result, while they may confront ethical issues, they are too abstract to address actual decision-making.


Adam Hammond 07.26.17 at 3:49 pm

Saving supplies is abstract. If the supplies are axiomatically linked to 100 lives (saved or not), then why not just have the 100 poor sick people tied to the tracks?


Neq_Neq 07.26.17 at 3:59 pm

Given that you are trying to build the problem around military intervention, shouldn’t the ‘normal’ train (10 vs 1) be a “Large Man” variant? Military intervention usually involves pushing someone onto the track.

Also, would it matter if the 10 people (normal train) were the last 10 people of thier respective racial/ethnic group? Just trying to figure out if genocide might be a special case that needs to be explicitly exempted.


Paul Kelly 07.26.17 at 4:15 pm

I’m not saying I disagree with the general sentiment of the post here, but for argument’s sake, maybe part of what Phil points out is that you can’t reduce such complex situations to oversimplified dilemmas. The group of people in poverty are probably divided in a number of ways and warring with each other; the ‘henchmen’ probably come from one subsection of a complex society and are engaged in some kind of large-scale oppression, maybe even genocide (in other words, it may not be possible to ensure an even distribution of aid, and it certainly isn’t possible to guarantee long-term effects); the cost of aid can’t be reduced to a one-off bill, but raises issues of sustainability (i.e. should systematic oppression be allowed to continue by outsourcing social welfare?); and, power systems, and their oppressive forces, can grow.

There are obviously a lot of different variables for different crises, so it seems a bit redundant to strip away the complexity. I wonder if it might be a bit logocentric to say, ‘no wars’. In any case, the argument certainly reminds me of the popular use of statistics as ‘facts’ in the news. The number of people is in some sense arbitrary. Why count hundreds of people that you can save with food versus ten bad guys? What do those numbers really mean? What might be the effect on society of killing a small number of oligarchs? Maybe it could destabilise global economic forces and make way for a more egalitarian politico-economic order. What would be the effect of killing me and a bunch of people from my town? Very little, I imagine.

Essentially, I don’t see the utility in dividing the options into two from a possible multitude of complex approaches. And there is a consequent danger of eliciting oversimplified motivations.

My apologies, if I’m missing the point. Certainly got me thinking, though.


Alex 07.26.17 at 4:27 pm

Perhaps the most moral choice is to do nothing at all. The second and third order effects of choosing to save a group of unique individuals with a network of connections vs one unique individual with a network of connections are impossible to predict. It’s not a black and white matter of the number of biological consciousnesses saved, as life vs death means very little in biology (see: Planet Earth and Planet Earth II).

For this binary choice, I’d say a human doesn’t have the knowledge needed to properly decide who gets to live and who gets to die, so it would be better to see how fate/reality/destiny/whatever we’d like to call actions that don’t involve our own conscious choices plays out rather than believing (with very little support or evidence) that we know what’s best. Either that or we should step onto the tracks ourselves in an attempt to slow the trolley even a tiny bit for whatever group might be hit, giving them a higher percentage chance of survival along with the knowledge that someone cared enough about them to sacrifice themselves. This would have a high chance of impacting their cultural beliefs and perhaps help them move forward with a new philosophy of helping other people at their own expense rather than simply choosing the greater over the lesser with no self-sacrifice requirement. So whichever person or group is saved would then have a higher likelihood of positively impacting the future, with our sacrifice as a founding pillar. In this scenario I could possibly see the case for switching the train to hit one person and then sacrificing yourself, but I still don’t think a philosophy of “the greater is more important than the lesser” is a good lesson to imbue.

For your version, I’d approach the scenario by asking what is the highest quality of life achievable for the largest number of people you have the power to impact?


Manta 07.26.17 at 7:10 pm

Your task is hopeless.
If someone takes seriously the “humanitarian” reasons given for wars, he is not going to listen to reason.


Joseph Brenner 07.27.17 at 12:19 am

Andre Mayer@5: “As a result, while they may confront ethical issues, they are too abstract to address actual decision-making.’

It’s supposed to be the other way around: an ethical thought experiment is supposed to clarify a general principle by creating a concrete situation that makes it clear. I agree that they rarely seem to work as intended– pretty frequently they end up inadvertently introducing side issues. They generally give you the feeling that the ideas under discussion aren’t applicable to the real world, or else you wouldn’t need to contrive bizarre hypotheticals to illustrate them.


Jeff R. 07.27.17 at 2:41 am

Mostly, ethical thought experiments are designed to drive a wedge between a student’s professed first principles and their actual position (usually on abortion, occasionally on war or public spending etc.)


CJColucci 07.27.17 at 3:17 pm

A variant that, if memory serves, comes from Bernard Williams.
You have been lost while hiking in the jungle in South America and come upon a clearing with a village. As you come closer, you see a dozen frightened peasants lined up in front of several armed soldiers, led by a Saturday morning western movie cliche dictator. You are captured and brought before him. After a brief interrogation, he is satisfied with your story and lets you go. He explains that the villagers had refused to pay their taxes and had been muttering about revolt and, therefore, he was going to execute a dozen of them to encourage the others. He is in a generous mood, though, and in honor of your escape from the dangers of the jungle he will, in your honor, shoot only one of them — but you have to choose. You can choose any way you like: seek volunteers, draw straws, or whatever, but you must choose. If not, he will have all twelve shot. Either way, you are free to go afterward and will not be harmed.What do you do?


Ebenezer Scrooge 07.28.17 at 2:10 am

Isn’t CJColluci@13 proposing Sophie’s Choice, with one extra kid?


RichardM 07.28.17 at 1:45 pm

Save the gangster’s victims, as that is time-critical. Organise a charity drive to replace the medical supplies. How is that even a question?

If you want to connect things, connect them, don’t just state they are connected, while setting up a scenario where they are not.

Pulling levers, if you understand what’s going on, can predictably divert trains. Pushing fat men doesn’t. Make the system of switches more complex, with parts hidden inside boxes, and you can gradually lose that understanding, piece by piece, until you know nothing. Having some of the opaque boxes labelled things like ‘imperialism’ or ‘terrorism’ can help, providing you trust the person who wrote the labels.


Eric Titus 07.28.17 at 7:14 pm

I think the issue with trolley problems in general is that they convince no one. People who disagree with you will recognize them as omitting key real-world details, people who agree with you will continue to agree with you.

Better to address the real world issue in all its frustrating complexity! Even opportunity costs, for all their issues, may influence a debate that frames these issues in economic rather than moral terms.


Collin Street 07.29.17 at 1:54 am

Mostly, ethical thought experiments are designed to drive a wedge between a student’s professed first principles and their actual position (usually on abortion, occasionally on war or public spending etc.)

As Richard M points out, real-world ethical situations are ones of incomplete knowledge; in that situation, heuristics, tendencies, and approximate solutions are the way forward.

Trolley problems strip away all that uncertainty and unclarity. It’s no surprise that our intuitive everyday understandings don’t work with trolley problems, but it’s precisely that that makes the “insights” gained from trolley problems meaningless for day-to-day real-world moral quandries.

So, yeah. Trolley problems drive a wedge between the situation as described and the moral framework of the students; the bits the wedge splits off are the problems that never arise in the real world. Keys and lampposts spring to mind.


Catchling 07.29.17 at 6:24 pm

Do you divert the first train, the second train, or neither?

Hopefully, most respondents will choose the second train.

Hang on. Setting aside my own views, I highly question this assumption that most people would choose any amount of medical supplies over even one human life. That’s what happens if you divert the second train, right? A single person gets killed?


Stephen 07.29.17 at 8:48 pm

CJ Colluchi@13: the heroic response is to say, sure, I’ll shoot one myself out of pure machismo; take the dictator’s pistol; shoot him.

Opinions may vary as to how far anyone of liberal, non-heroic principles could be expected to do that.


Petter Sjölund 07.30.17 at 6:32 am

As Richard M points out, real-world ethical situations are ones of incomplete knowledge; in that situation, heuristics, tendencies, and approximate solutions are the way forward.

I guess one purpose of thought experiments is to lay bare those ”heuristics, tendencies, and approximate solutions” and how they are generally applied. That can be a very useful way to detect things like bias and prejudice. One just has to be aware that perfectly adequate general principles may seem absurd or to give absurd results once we are detached from reality.


Dmichaele 07.30.17 at 7:12 pm

Hypotheticals, especially those with binary choices, are not rooted in reality and therefore offer no actual solutions to problems of war, poverty, injustice, etc.

ISIS, for instance, is real. They are not statesmen, diplomats, negotiators or consensus builders. They even consider Hamas to be “infidels” because Hamas is seeking legitimacy via the United Nations—a decidedly Western invention. They ask, “Why seek solutions from Ban Ki Moon or Antonio Guterres when Allah is the source of truth and power?” Meanwhile their genocidal actions continue. It wasn’t until their murderous action was met with a more substantial force that they began to lose their advantages. Is it morally wrong to fight them off militarily?

President Obama believed that releasing terrorists from Gitmo would mitigate resentment toward the USA and thus deter ISIS’s recruiting abilities. As several journalists have pointed out, members of ISIS were, in fact, praying for the release of the Gitmo prisoners and they viewed the prisoner’s release from Gitmo as Allah answering their prayers. Rather deterring ISIS’s recruiting, it emboldened them by allowing them to claim that Allah was listening to their prayers, thereby legitimizing their quest to establish their Caliphate and getting more fighters in their ranks.

Again, there are many factors that binary hypotheticals do not address.


CJColucci 07.31.17 at 3:00 pm

The dictator may be vicious, as well as possessed of a bizarre sense of humor, but he isn’t stupid. He’s not going to give the hiker a gun, or risk his taking one. Presumably, the hiker has already figured that any attempt at a heroic response would be futile.


Chris Borthwick 08.01.17 at 4:49 am

If we are taking on essentially all-knowingness about consequences, let’s switch to the more venerable problem from Voltaire’s Zadig:
<b<Zadig, tho’ astonish’d to the last Degree, attended him to their last Stage, which was to the Cottage of a very virtuous and well-dispos’d Widow, who had a Nephew of about fourteen Years of Age. He was a hopeful Youth, and the Darling of her Heart. She entertain’d her two Guests with the best Provisions her little House afforded. In the Morning she order’d her Nephew to attend them to an adjacent Bridge, which, having been broken down some few Days before, render’d the Passage dangerous to Strangers.
The Lad, being very attentive to wait on them, went formost. When they were got upon the Bridge; come hither, my pretty Boy, said the Hermit, I must give your Aunt some small Token of my Respect for her last Night’s Favours. Upon that, he twisted his Fingers in the Hair of his Head, and threw him, very calmly, into the River. Down went the little Lad; he came up once again to the Surface of the Water; but was soon lost in the rapid Stream. O thou Monster! thou worst of Villains, cry’d Zadig! Didn’t you promise, said the Hermit, to view my Conduct with Patience? Know then, that had that Boy liv’d but one Year longer, he would have murder’d his Foster–Mother. Who told you so, you barbarous Wretch, said Zadig? And when did you read that inhuman Event in your Black–Book of Fate? Who gave you Permission pray, to drown so innocent a Youth, that had never disoblig’d you?
No sooner had our young Babylonian ceas’d his severe Reflections, but he perceiv’d that the old Hermit’s long Beard grew shorter and shorter; that the Furrows in his Face began to fill up, and that his Cheeks glow’d with a Rose-coloured Red, as if he had been in the Bloom of Fifteen. His Mantle was vanish’d at once; and on his Shoulders, which were before cover’d, appear’d four angelic Wings, each refulgent as the Sun. O thou Messenger of Heaven! O thou angelic Form! cry’d Zadig, and fell prostrate at his Feet; thou art descended from the Empireum, I find, to instruct such a poor frail Mortal as I am, how to submit to the Mysteries of Fate. Mankind in general, said the Angel Jesrad, judge of the Whole, by only viewing the hither Link of the Chain. Zadig, begg’d Leave to speak. I am somewhat diffident of myself, ’tis true; but may I presume, Sir, to beg the Solution of one Scruple? Would it not have been better to have chastiz’d the Lad, and by that Means reform’d him, than to have cut him off thus unprepar’d in a Moment. Jesrad, replied, had he been virtuous, and had he liv’d, ’twas his Fate not only to be murder’d himself, but his Wife, whom he would afterwards have married, and the little Infant, that was to have been the Pledge of their mutual Affection.
So, do you drown the kid?

Can’t resist throwing in, so to speak, another pretty boy;


Dave 08.01.17 at 7:52 pm

I agree that this is silly. For all we know, the ten people the gangsters would have killed were also train directors guiding medical supplies. A hundred deaths is a tragedy, but targeted killings often do far more damage than random death.

Another consideration. How do the families of those ten people the gangsters killed react? The families of the sick may lay blame and plot revenge, but they likely won’t if they think the disease is “no one’s fault.” However, where there is murder involved, there’s also the desire for vengeance. If the murder victims’ families kill a bunch of gangsters and then the gangsters’ families start killing them, then you’ve started a cycle of violence that could leave thousands dead instead of just a hundred.

Of course, the counterpoint to that is that, if you kill the gangsters, then their families might go after you so, if that’s something that concerns you, it’s better to save the medical supplies and tell yourself you’re doing the morally correct thing anyway, so you weren’t being selfish. Then you can feel virtuous and stay safe and who cares about the ten people the gangsters would have killed? There’s a twisted logic of self-preservation at the core of anti-interventionism that doesn’t necessarily lead to unethical choices (after all, your family being roped into the cycle of violence has a pretty similar utilitarian cost to it happening to a bunch of strangers), but is certainly self-serving.

The worst option, of course, is the one a lot of real world anti-interventionists favor, which is to invoke the sick village as an excuse to do nothing about the gangsters, and then do nothing to help the sick village, either. That’s basically what “America First” means, for instance.


Dave 08.01.17 at 8:23 pm

The other issue with trolley problems is they fail to take into account the power dynamics of being a switch operator. Obviously a venal, cowardly, or selfish person is going to choose the option that has the best outcome for them personally. In this example, perhaps the gangsters can offer a much bigger bribe than the poor villagers, or, perhaps, the switch operator is motivated by the praise and sense of virtue they’ll get if they save the villagers. This is why trolley problems usually assume a perfectly rational and disinterested switch operator. Perfect objectivity is, of course, impossible, but, even if we accept that premise, there’s still an issue.

Let’s assume two things. First, that the switch operator knows that they’re the best switch operator and no one else can do the job as well as they can. Second, let’s assume that the switch operator is employed by a committee that will fire them if they ever disagree with the choice the switch operator makes. Let’s say that the committee will almost always agree with the switch operator about which switch to pull, so avoiding getting fired and replaced by a less competent switch operator is always better than getting fired.

So, one day, our trolley operator sees that that there are ten people on one track and the committee chairman’s son on the other. Ten people die. That’s the problem with being a trolley operator in a democracy.

Comments on this entry are closed.