Suppose you see a trolley racing down its tracks, unable to stop. Further down, there are five people tied to said tracks. You, meanwhile, are standing next to a lever that will shift the trolley’s course towards a side track before it runs over the five people. On that track, however, one other person is tied.
What do you do? And why?
The Trolley Problem in Pop Culture
This “trolley problem” entered recent pop culture in an episode of Unbreakable Kimmy Schmidt (May 2017), where the titular character is presented with the experiment in a philosophy course at Columbia. A few life epiphanies later, Kimmy finds herself taking a qualifying test to be a crossing guard and must act out her solution to the problem when a cardboard cutout of a forward-moving truck (which is, in its own way, “unbrakeable”) races towards an intersection with cardboard cutouts of five children down one path and a businessman down the other. Kimmy’s solution is to motion the truck forward into herself, thus saving both the five children and the businessman.
Meanwhile, there’s The Good Place, an overtly philosophical sitcom that addresses ideas like utilitarianism, Kant’s categorical imperative, Kierkegaard’s leap of faith, the concept of the self, and the trolley problem. With its afterlife setting and divine-like beings, The Good Place has protagonists simulate the trolley problem in lifelike detail multiple times.
A few episodes later, a real-life choice confronts one character―and he realizes that it’s the trolley problem. His solution is the same as Kimmy’s: He sacrifices himself, letting demons apprehend him in “The Bad Place,” so that his friends can all escape.
It’s intriguing enough that both sitcoms mention this thought experiment, let alone that they both answered it the same way. Their solution is a testament to the esteem we have for self-sacrifice . . . but it’s also cheating. That option isn’t actually on the table. More importantly, there is a real lesson we can draw from the actual right answer―the answer that explains, in as few general principles as possible, directions that are intellectually satisfying and hold up under numerous different scenarios.
The Fat Man Variant
Behold the trolley again, racing before your eyes towards five random people. This time you don’t have a lever, but you are on a bridge over the tracks, and beside you, spectating from the very edge of the railing-free bridge, stands a very fat man. It occurs to you that while you may be ninety pounds soaking wet, you could give your bystander one unsuspecting shove, and his landing in the trolley’s path would obstruct its trajectory, sacrificing one life to save five.
What do you do?
When you were asked earlier, you probably thought you should pull the lever and save the five. Most people do―by survey data, something like 90% of people say they would, intuiting that that’s the right thing to do even if they can’t say why. When many people do give a justification, it’s often the utilitarian reason: five is more than one. Maximizing happiness for the most people means sacrificing the one.
When you present the fat man variant, however, people balk. This seems like a worse thing to do, so people try to distinguish the two scenarios. From a utilitarian perspective, however, nothing has changed. One life or five? That is all.
These two cases are useful for exposing one of the embarrassing features of utilitarianism as a moral principle: A utilitarian framework might morally require anything of you, however reprehensible, in some scenario. No action is absolutely off-limits.
As The Good Place’s Jason Mendoza says in Season 1, “It’s like, I knew this girl Sheila . . . Sheila was gonna get married to my boy, Donkey Doug, and make him move to Sarasota. It would’ve broken up my whole break dancing crew, and Donkey Doug was our best pop-and-locker. So I hid a bunch of stolen boogie boards in Sheila’s garage and called the cops. I framed one innocent [person] to save a 60-person dance crew.” Framing people is wrong, but utilitarianism says that if that’s the way to render a benefit to 60, maybe it’s called for.
Many examples, in fact, highlight the shortcomings of utilitarianism. Meanwhile, the trolley problem leaves us wanting more: What is the right thing to do, and why?
The Principle of Double Effect
The answer actually appears in The Good Place (where it gets very short shrift), in the episode “Derek.” The right framework, formulated by Thomas Aquinas, is the principle of double effect. This is a way of analyzing actions by looking at different types of effects they have.
The principle’s first component is the idea that you may not perform an action that is itself morally wrong. This includes theft, murder, adultery/incest/fornication/rape, and shoving fat men off bridges. It is this instinctual moral knowledge that causes us to balk at the fat man example. Similar setups are a common variant of the trolley problem in fiction, played to the hilt for emotional torment.
Returning to the initial trolley problem, we see that this lens reveals differences that a utilitarian one does not: Pulling a lever means directing the trolley’s path away from one set of tracks. That diversion, being morally neutral in itself, is a different sort of action from pushing a fat man to his doom. The action of pulling the lever is a different sort of action from sending a fat man into the trolley’s path, but after diverting the trolley we still care about where it goes, and it’s barrelling towards someone else.
This is the second, or “double,” effect of your pulling the lever. As Aquinas puts it, “Nothing hinders one act from having two effects, only one of which is intended, while the other is beside the intention” (Summa Theologica II-II, Qu. 64, Art.7). The second criterion of the principle of double effect is that the bad effect is a side effect–at least, it must not be the instrument for the good effect. The one person getting run over is a secondary effect since his being there or not isn’t essential to diverting the train, whereas the other setup requires the fat man to perish to stop the trolley and achieve the good objective. Bad side effects may be permissible in the same way that dreamt-of ends do not justify wicked means.
Then there is a third component. Suppose that the trolley is barrelling towards one innocent man, and the alternate track has five people tied down to it. Switching is not called for in this case because the secondary/double effect must be proportional in relation to the harm or evil you’re thwarting.
The Final Component
The final component to this principle is subtler but no less necessary. Imagine that instead of a random fat man, your bridge bystander is someone very specific―maybe an opposition pick for a Supreme Court swing seat, or your own mother-in-law, or someone like that. Someone who isn’t Hitler-level evil per se but whom you’d like to just be gone. Suddenly the trolley problem might look like a golden opportunity―you get to heroically save five people and be rid of a nuisance.
Thomas Aquinas tells us, as you might have guessed, that that perspective is not OK. An action becomes wrong if you’re doing it for the purpose of a secondary result that is in itself wrong. The purpose of your action must be something affirmatively good, such as saving five lives, for any secondary harm to even qualify as maybe-allowable.
That last criterion is different from the other three in that it touches on the morality of the choice and the disposition of your heart. Recall Jesus saying in the Sermon on the Mount that “everyone who looks at a woman with lust for her has already committed adultery with her in his heart” (Matthew 5:28 NASB). That particular action (looking) is wrong, but only because of the disposition that the actor has―it’s not a generally wrong action that we should, say, legislate against (laws are, by definition, general rules of conduct).
To fully understand the trolley problem, and what makes those choices right or wrong, we need this subjective component, too.
Altogether the principle is this:
You may perform an action that is itself morally neutral, for the purpose of directly effecting some good, even though some evil, which is not disproportionate, also results as a secondary effect.
That is the full answer to the trolley problem. The problem has spawned other variants as well (such as the “fat villain,” where the fat man is no innocent bystander but rather the architect of the scenario), but many raise far different sorts of questions or the application of the principle of double effect effortlessly resolves them. Sometimes, of course, using double effect means contradicting our own intuitions―some people, after all, would shove the fat bystander off the bridge. More often than not it is a way of clarifying and articulating intuitions with which our consciences already supply us.
Using double effect pays dividends in real life when dilemmas confront us and we need to take actions laden with undesirable consequences. While The Good Place may not have fully presented the answer, the show at least had the vision and audacity to raise such a hypothetical to popular consciousness. Even though it is now concluded, its participation in the ongoing conversation of these great ideas will be long felt, and that is a good thing.
Real World Applications
Maybe, as diverting as this is, it all sounds pointlessly abstract to you. If so, rest assured that the principle that this hypothetical draws into focus often applies to the real world. Claims about real-life runaway trolleys might be greatly exaggerated, but the principle applies to superficially different dilemmas.
Some reports out of China suggest that they have arrived at an answer to one variant of the thought experiment–you are a surgeon with five men dying of different organ failures, and in waltzes a hypochondriac in perfect health whose organs, harvested and distributed, would save five lives if only his own life were forfeit. Even in America, some medical questions present conundra that this principle helps resolve–suppose a terminally-ill patient is in severe pain, and the only drugs you have to relieve the pain would, as a side effect, probably shut down certain vital systems and trigger death. Is that different from euthanasia? I think that Thomas Aquinas can explain why for us.
This also applies when people plead “the life of the mother” as an exception for abortion. So far as I am aware, the valid instance of this is an ectopic pregnancy. Is it inconsistent for pro-lifers to admit this as an exception?
Philosophy professor Nathan Schlueter (from whom I myself first learned of the principle) applies the principle of double effect in First Things thus: “The classic case here is ectopic pregnancy, in which the fallopian tube must be removed, even when the result will be the death of the child [emphasis mine] . . . The death of the child (a great material evil) is tolerated (not directly intended or caused) because the moral object (the life of the mother) is sufficiently proportionate to it [emphasis his].” This emergency procedure is a different action from the actions grouped under the euphemism of “abortion,” whether that’s chemical burning or physical butchering.
There are non-medical examples too. Engineers (and governments) are even now debating dilemmas much like the trolley problem concerning how to program self-driving cars if they find themselves in such a situation. Similarly, much lawmaking would benefit if legislators examined outcomes using double effect instead of the tired and flawed utilitarian lens. If you keep your eyes open, you’ll realize that any choice that troubles you due to consequences conflicting with apparently wrong deeds might surrender clarity if you viewed it with the principle of double effect in mind.
If none of these scenarios feel relevant enough for you, I suggest you enjoy this video where someone fabricated the exact scenario for some unsuspecting subjects and filmed them resolving it for themselves. You never know when a philosophical thought experiment might confront you and demand a resolution.