Monday, May 12, 2014

Rethinking the Trolley Problem

There is a well known thought experiment called The Trolley Problem. You may have come across it before. A basic description of the problem is:
There is a runaway trolley barreling down the railway tracks. Ahead, on the tracks, there are five people tied up and unable to move. The trolley is headed straight for them. You are standing some distance off in the train yard, next to a lever. If you pull this lever, the trolley will switch to a different set of tracks. Unfortunately, you notice that there is one person on the side track. You do not have the ability to operate the lever in a way that would cause the trolley to derail without loss of life (for example, holding the lever in an intermediate position so that the trolley goes between the two sets of tracks, or pulling the lever after the front wheels pass the switch, but before the rear wheels do). You have two options: (1) Do nothing, and the trolley kills the five people on the main track. (2) Pull the lever, diverting the trolley onto the side track where it will kill one person. Which is the correct choice?
(Source: Wikipedia)

Most people tend to conclude that the right course of action is to pull the lever and kill the one person rather than the five. I also agreed with this until thinking about it some more recently, and I now think that the right choice is probably to not touch the switch and let the five people die. In the remainder of this post I will try to back up my choice and either convince you that it is the right one, or at least give you some good food for thought and make you examine your own moral intuitions a bit.


The first thing to point out is that when most people justify why they would pull the switch, they tend to appeal to some form of utilitarianism, that is, given two bad choices, the right one is the one that minimizes death/suffering. This seems fairly sound. After all, if your choice is one person dying or five people dying, why wouldn't you save as many lives as possible?

However, a variant on this problem shows that the reasoning is probably more complex than this. The variant, sometimes called The Fat Man, is a bit contrived and I'm not a big fan of it, but it's the standard one so I'll stick with it:
As before, a trolley is hurtling down a track towards five people. You are on a bridge under which it will pass, and you can stop it by dropping a heavy weight in front of it. As it happens, there is a very fat man next to you – your only way to stop the trolley is to push him over the bridge and onto the track, killing him to save five. Should you proceed?
(Source: Wikipedia)  

Now, it does require a certain amount of leeway to accept that you could be in this situation and somehow be certain that pushing the fat man onto the tracks would definitely stop the trolley, and that nothing else within reach will do.

But, accepting the premises, most people in this case will not support pushing the man onto the tracks, even though the utilitarian argument should still apply in this case. Is it the direct action that puts people off? If the fat man was instead standing on a trap door and you could drop him onto the tracks with a lever, would that make it easier for some people? Or is it the feeling that you're directly causing the death of the fat man, while in the original case the death is an indirect result of your actions?

To be clear, just as I thought the lever should not be pulled in the first example, I think that the fat man should not be pushed in this example.

Reframing the problem

Sometimes clarity can be obtained by reframing a problem and looking at it in a different light. Since this is just a thought experiment, we can take the basic dilemma and set it up using different elements, which has been done in various ways by other people over the years.

One version is The Crashing Plane: a plane has critically malfunctioned and is going down over a populated area. The pilot cannot stop the crash, but he has some limited choice in steering the plane towards a more populated area or a less populated area. He can safely assume that his choice will result in a lot of deaths in the former and less deaths in the latter, but people will die either way.

Most people would say that the pilot should aim for the less densely populated area, and I agree with that choice. Does this contradict my stance on the two previous examples? I don't believe so, and I will explain why shortly.

But first, here is another version of the problem, and the one that for me framed it in a way that made me change my mind about the original trolley problem. This one is The Transplant Problem:
A brilliant transplant surgeon has five patients, each in need of a different organ, each of whom will die without that organ. Unfortunately, there are no organs available to perform any of these five transplant operations. A healthy young traveler, just passing through the city the doctor works in, comes in for a routine checkup. In the course of doing the checkup, the doctor discovers that his organs are compatible with all five of his dying patients. Suppose further that if the young man were to disappear, no one would suspect the doctor.
(Source: Wikipedia)
In this case, I would say that it is wrong to kill the one person to save the lives of the other five. An interesting question for this version is whether or not it makes a difference that that the doctor would not be suspected of committing the murder. The purpose of this clause is probably just to simplify the problem to not have any further repercussions on the doctor, so the choice becomes simply about the lives of the five patients and the one traveller. But in this clause I think also lies the key to the whole problem.

Wider consequences

What if we were to take The Transplant Problem, but remove the secrecy aspect of it? What if people knew that the doctor had made this choice? In all of the previous versions of the problem there was no requirement of secrecy. To be a proper solution to a moral problem, I think that most people would agree that the choice should either be defendable in court, or if the law doesn't currently support it, then it should be modified to make the choice legal. After all, does it really make sense to say, for example, that the right moral choice is to push the fat man off the bridge, but that if you did it, you should go to jail? Morals and laws might get out of sync in practice, but does it make sense to want them to be out of sync in principle? (maybe there is a case to be made for this, and if anyone can think of one, I'd love to hear it!)

So for The Transplant Problem, we need to ask: would we support the law changing in such a way that doctors would be allowed to kill one of their patients if it meant that they could save others? Would you want to live in a society where every visit to your local GP could potentially result in you having your organs harvested, all in the name of the greater good?

I think very few people would support this. And so the key to all of these problems is that our moral intuitions can get pushed in a certain direction when we are only thinking of them in an isolated context, but we can't do that. We need to think about these choices in a broader context where other people might be put in the same position of having to make the choice, and you or people you care about are potentially on the receiving end of those choices. In this context, things can look very different.

Let's now backtrack through the versions of the problem and see if I can now justify my choices:
  • The Transplant Problem: As stated, I don't think many people would want to live in a world where doctors harvest the organs of people against their will to save other people's lives. The resulting fear and paranoia, not to mention the obscene violation of an individual's right to be secure in person, would far outweigh the lives saved. So it also shouldn't be okay for a doctor to do this in the case where it was illegal but he didn't get caught.
  • The Crashing Plane: In this case, the pilot is an active participant in the scenario. He has created the very threat that is going to cause deaths one way or the other, and his only choice is whether to minimize the number of deaths, so steering towards the lesser populated area is the right thing to do.
  • The Fat Man: In this case, you are choosing to sacrifice one person's life to save the lives of others. But by what right do you get to decide that someone should die to save other people? If you were the fat man, would you want to live in a society where other people would decide to kill you for the greater good?
  • The Trolley Problem: Whether you push a person to their death or pull a lever knowing that a person will die as a result, we don't have the right to choose to sacrifice one person's life to save others. Unless our actions have already put people's lives in peril and we're simply trying to minimize the damage (as in the crashing plane scenario), we need to respect the right each individual has to be secure in person.
Now, I have to admit that I still have trouble feeling that my choice for the trolley problem is right (and even the fat man, to some degree), but the transplant problem seems much clearer and more obvious to me, and since I can't see a logical difference between them (as opposed to the crashing plane, which I think is different), I have to conclude that it is the correct choice even if my ape brain doesn't like it.


I thought of something that might make the choice in the trolley problem a little easier to understand, which is illustrated nicely by, of all things, what Batman says to Ra's Al Ghul at the end of Batman Begins: I won't kill you, but I don't have to save you.

There is a moral difference between killing a person and simply not acting to save them, and this is a key part of the trolley problem, where pulling the lever is actively killing a person, while not touching it is not saving five people. While I think that the former is worse than the latter, I'm not sure that this would be true for any number of people. If it was the choice between actively killing one person or failing to save the lives of 1 million people, would the choice be the same. And if not, where is the dividing line? Can we justify having one? This may be a bit like the case for abortion, where most people (who support abortion at all) think that it's somewhere between conception and birth, but actually placing a specific cutoff point seems arbitrary. Or maybe this situation is different, and we can't ever really justify killing one person as the better choice?


  1. I'm not going to say whether this is right or wrong (after all, they're thought experiments - there isn't really a right or wrong), but I do have a bit of an issue with some of the logic used. I've included my thoughts below:

    The Transplant Problem: As you say, one of the issues with this one goes beyond the individual circumstances. It has consequences of massively negatively impacting society as a whole through fear and so on. This makes it difficult to apply a direct Utilitarian argument to it.

    Transplant problem + The fat man: In both these instances, the potential single victim either has an opportunity or could be given an opportunity to make the decision themselves to sacrifice themselves - in which case it could be left up to them. In my opinion, this makes it a bit different. I think it's also difficult to imagine the Fat Man problem properly for the simple reason that it's ridiculous - there is obviously no way a fat man will stop a trolley - so for me, even trying to imagine the situation makes me just feel like I'm killing an extra person, not saving five.

    The Crashing Plane - You say that the pilot created the threat in the first place. But that may not be true. What if it was entirely a malfunction by the plane? Or what if a passenger took over the plane after the pilot passed out or something. I would say the passenger should still choose a less populated area to crash if possible, even though he's just entered the situation.

    The Trolley Problem: The problem here is the snap decision that needs to be made - in the others there is a bit more time. I think that potentially changes the "correct" answer. We are allowed to have time to think about and come up with the best answer. In that situation you would have to go with your gut feeling, and you appear to imply that your gut would say pull the lever.

    I think my issue with the overall logic is that you've taken a very small sample of potential scenarios, found a link between a couple and decided that applies in all instances (though even amongst those, you suggest one is still different). But I don't think it's that simple. The circumstances of the choices DO make a difference.

    For example, take another example for a surgeon: He is in the middle of an operation on one person. The operation is certain to save the persons life, but will take quite a while. He is told that he can save 5 other people with a quick procedure, which he must go and start now, and he is the only person that can do it. But if he leaves the current operation, that patient will die. Personally that makes it a tough choice, but in a real situation (say, an army hospital - in fact I've kind of stolen this off M*A*S*H) the surgeon should save the other lives. Now let's complicate that by saying that to save the THREE of the other lives he needs organs from that one person. We're getting closer to the Transplant Problem, and the waters are much muddier now.

    I also feel that "I won't kill you, but I don't have to save you." is an oversimplification. If you can easily save an innocent person, with no risk to yourself or others (and no real negative consequences to yourself, such as loss of money or something) then I think there is very little, if any, moral difference between letting them die and killing them. e.g. With the trolley problem. Remove the one person from the other track. Can you still argue that letting the Trolley continue is just "not saving them"? I don't think so, I think you're just about actively killing them.

    So my overall point is that I don't feel you can take the logic from one and apply it to all.

    Note that personally I feel that either option is correct in the original trolley problem, because neither option is right. I think most people would have trouble fronting up to the families of those that died regardless of whether you pulled the lever or not.

  2. I think you make a lot of excellent points here, and I should make it clear that I'm far from convinced by my own logic! :)

    While I feel that there is a certain logical thread that connects all of these problems, I agree that the circumstances do seem to make a difference, and as you point out in some of your additional examples, you can make modifications to the original problems that have a good chance of changing a person's stance on the problems.

    I identified the crashing plane as feeling different to the other three, but I can easily see small changes that would make me rethink this. I should be clear that in this scenario I wasn't suggesting that the pilot was responsible because he made a mistake (I actually assumed a plane malfunction), but that since he was at the controls of the thing that was causing the threat, it automatically made him 'involved' rather than a bystander, so there doesn't seem to be a 'do nothing' option like in the other cases. But I'm not sure that this is actually sound.

    Like you pointed out, what if instead a passenger walked into the cockpit and had to make a choice. I agree that he should also aim for the less populated area. But is this really different to the bystander with the lever? Or how about we take the original problem, but make the lever be inside the trolley, and the driver is making a choice whether to switch tracks? Or what if a passenger on the trolley walks into the driver's cabin and has to make the choice. Now we seem to have blurred the lines between the trolley problem and the crashing plane.

    The case of the surgeon having to pick who to operate on is a bit different I think, because in this situation all people will die if he does nothing, so it's not a question of explicitly killing a person to save another, but one of not being able to save everyone, so having to triage, which feels more justified. But then you suggest the alternative of whether to take the organs of one dying person to save several other dying people. This one I think is different to the previous one because we've switched from picking who to save to explicitly killing one person to save others, i.e. the latter case requires him to break the Hippocratic Oath, while the former one does not. But then what if he can wait for the one guy to die and then take his organs to save the others. He didn't kill him because he was dying anyway.

    And now, for all of these scenarios, what if we have some fun and start introducing loved ones? So we have the trolley problem with 5 strangers on one track, and your child/partner/sibling/parent on the other? How many people would let their loved one die? What if it was 10 strangers? 100? Here our moral logic is clearly swayed by our emotions. And I wonder how much this also ends up affecting the other scenarios to various degrees.

    In the end, I suspect that no matter what stance people have on these original problems, there are infinite little variations you can make that will eventually reach a point where a person changes their mind, but probably can't logically justify it. I think this might be another case of the kind of moral paradox I discussed in my previous blog post about luck and morality, where our intuitions and sense of justice have been shaped by evolution to solve certain problems, but they're just heuristics with no guarantee that we can create a bulletproof moral theory that has a logic answer for every situation, and the kinds of discussions we have here make this clear in the same way that optical illusions reveal the gaps in our visual processing caused by the evolved visual heuristics we have.