The Trolley Problem — Would You Sacrifice One to Save Five?

Opening: You, the Lever, and a Choice
The First Twist: The Large Man
Two Great Ethical Theories: Look at Results, or Look at Principles?
- Utilitarianism: The Greatest Happiness for the Greatest Number
- Deontology: Some Things Must Not Be Done
The Doctrine of Double Effect: The Line Between Intention and Side Effect
A Lattice of Variations: A Laboratory for Intuition
What Neuroscience Saw in Moral Intuition
From Thought Experiment to Reality: The Ethics of Self-Driving Cars
Many Perspectives, and Open Questions
Closing: Food for Thought
References

Opening: You, the Lever, and a Choice

Picture this. You are standing beside a train track. Far down the line, a trolley with failed brakes is barreling toward you at terrifying speed. On the track ahead, five workers are absorbed in their labor. They have not noticed the oncoming trolley, and there is no time to escape. Left as it is, all five will die.

But beside your hand is a lever. Pull it, and the trolley will switch onto a side track. The trouble is that one worker stands on that side track too. If you pull the lever, the five live, but that one person dies.

So, will you pull the lever?

This is the famous "Trolley Problem." It was first posed in 1967 by the British philosopher Philippa Foot, and later refined with great precision by Judith Jarvis Thomson. On first hearing, it sounds like simple arithmetic. Five lives against one, surely we save the five? And yet this thought experiment has tormented philosophers for more than half a century, and today it has psychologists, neuroscientists, and the engineers building self-driving cars all scratching their heads.

In this essay we will follow why the trolley problem is so stubbornly hard, what two great ethical theories it exposes, how our brains process the dilemma, and why this old puzzle has become one of the most practical technological questions of the twenty-first century. Let me say upfront: this essay offers no "correct answer." Its goal is to help you see your own intuitions more clearly.

The First Twist: The Large Man

Most people, faced with the original lever scenario, say they would pull it. Sacrificing one to save five feels rational. But Thomson's second version changes the story.

This time there is no lever. You stand on a footbridge spanning the track. Below, the trolley races toward the five. Beside you, leaning on the railing, is a stranger of considerable size. If you push him onto the track, his large body will stop the trolley and save the five. He, of course, will die.

So I ask again. Will you push him?

Remarkably, most people answer "no" to this one. The outcome is identical: one dies, five live. Arithmetically it is exactly the same as the first situation. And yet our intuition reacts in the opposite direction. Pulling a lever feels acceptable; pushing a person feels horrifying.

Why? This very mismatch of intuitions is the heart of the trolley problem. Within us seem to live two conflicting moral voices. One says, "Count the numbers, save more lives." The other says, "Do not use a person as a tool; do not harm directly."

Two Great Ethical Theories: Look at Results, or Look at Principles?

These two voices in fact represent two great traditions of moral philosophy.

Utilitarianism: The Greatest Happiness for the Greatest Number

The first voice, "save the five," aligns with consequentialism, and especially its flagship, utilitarianism. Shaped by the eighteenth-century Englishman Jeremy Bentham and the nineteenth-century John Stuart Mill, this view holds that the morality of an act is judged by the results it produces. The right choice is the one that brings greater happiness (or less suffering) to more people.

From the utilitarian point of view, the answer is clear. Five lives carry more value than one, so whether you pull a lever or push a person, saving the five is right. If the result is the same, the method does not matter.

Deontology: Some Things Must Not Be Done

The second voice, "do not push the person," aligns with deontology. The German philosopher Immanuel Kant is its leading figure. Deontology holds that the rightness or wrongness of an act is determined not by results but by the nature of the act itself and the moral principles (duties) we must follow.

One of Kant's central maxims is to "treat humanity never merely as a means, but always at the same time as an end." Pushing a man off a footbridge uses his body as a "tool" to stop the trolley. It treats him not as a person but as a brake pad. For the deontologist, this is a line that must not be crossed, no matter how good the outcome.

Below is a brief comparison of the two positions.

Aspect	Utilitarianism (Consequentialism)	Deontology
Standard of judgment	The result of the act	The act itself and its principle
Central question	Which produces more happiness	Does this act conform to moral law
Key figures	Bentham, Mill	Kant
Trolley answer	Save the five (pull / push)	Do not use a person as a means (do not push)
Strength	Clear and calculable	Protects individual dignity
Weakness	May justify sacrificing the few	Can yield rigid conclusions

The Doctrine of Double Effect: The Line Between Intention and Side Effect

Is there a more refined way to explain why pulling the lever is permissible but pushing a person is not? Here enters the doctrine of double effect, which traces back to the medieval theologian Thomas Aquinas.

According to this principle, when an act produces both a good and a bad result at once, it can be morally permissible if certain conditions hold: the bad result must be a foreseen side effect rather than the intended aim, and the good result must not be achieved by using the bad result as a means.

Apply this to the lever case. Your intention in pulling the lever is to save the five. The death of the one on the side track is an unintended but foreseen side effect. Had that person not been there, your plan would have succeeded just the same. His death is not "required" to save the five.

The footbridge case is different. There you use the man's death (more precisely, his body blocking the trolley) as a means to save the five. If he survives, your plan fails. His death becomes part of your intention. This very difference may be the key that divides our intuitions about the two cases.

Of course, the doctrine of double effect draws criticism too, because the boundary between "intention" and "side effect" is not always clear. To test it, Thomson devised yet another variation.

A Lattice of Variations: A Laboratory for Intuition

The charm of the trolley problem is that its endless variations let us dissect our moral intuitions with precision. Let us look at a few more.

The Loop: Suppose the side track curves back and rejoins the main line in a loop. On that loop stands one person whose body is heavy enough to stop the trolley. If you pull the lever, the trolley travels the loop, strikes that person, and halts. The five live. But here the one person's body serves as the "means" of stopping the trolley. It feels like the lever case, yet structurally it resembles the footbridge case. This subtlety exposes the limits of the doctrine of double effect.

The Doctor with the Drugs: A doctor has five dying patients, each needing a different organ. A healthy person happens to come in for a checkup. Harvesting his organs would save all five. Should the doctor do it? By utilitarian arithmetic this is identical to the lever case. Yet almost everyone answers, "Absolutely not." This shows that mere number-crunching cannot explain our moral intuitions.

These variations, like a chemist swapping reagents to observe a reaction, reveal which variables our moral responses are sensitive to. Whether we make direct contact, whether we use a person as a means, whether we physically push or merely flip a switch — all of these subtly shift our judgment.

What Neuroscience Saw in Moral Intuition

In the twenty-first century, a new tool entered this philosophical riddle: brain imaging. The American psychologist and philosopher Joshua Greene conducted a landmark experiment with colleagues in 2001. They presented people with moral dilemmas like the lever and footbridge versions while observing brain activity with fMRI (functional magnetic resonance imaging).

The results were intriguing. When people considered dilemmas involving "direct and personal" harm, like the footbridge case, brain regions associated with emotional processing (the medial prefrontal cortex, the amygdala, and others) became more active. By contrast, in "impersonal" dilemmas like the lever case, regions handling reasoning and calculation (such as the dorsolateral prefrontal cortex) were relatively more engaged.

On this basis Greene proposed a "dual-process theory." Within us live two systems. One is a fast, automatic emotional response that generates intuitions like "do not directly harm a person." The other is a slow, deliberate rational calculation that issues utilitarian judgments like "save more lives." The image of pushing a man off a bridge rings a loud emotional alarm; pulling a distant lever does not.

A note of caution is needed here, though. The fact that a brain region activates does not in itself prove that a judgment is "right" or "wrong." An intuition born of emotion is not necessarily mistaken, nor is rational calculation necessarily correct. Neuroscience can show us how we judge, but it cannot decide how we ought to judge. Greene's own interpretation continues to be vigorously debated in academic circles.

From Thought Experiment to Reality: The Ethics of Self-Driving Cars

For a long time the trolley problem was treated as an entertaining classroom brain-teaser. After all, you almost never find yourself standing beside an actual track lever. Then, in the twenty-first century, this thought experiment suddenly became an urgent practical problem, thanks to the arrival of self-driving cars.

Imagine an autonomous vehicle. One day, as the car travels down a road, several pedestrians suddenly dart into its path. They are too close to stop with the brakes. If the car continues straight, the pedestrians are hurt. If it swerves, the pedestrians are spared, but the car strikes a wall and endangers its occupants. How should the car be programmed to behave?

This is no longer a hypothetical. Engineers must in fact decide in code, in advance, what the car will do in such situations. The trolley problem has migrated from an abstract lever to a concrete algorithm.

To put this question to the public, MIT's Media Lab ran an online experiment called the "Moral Machine." Millions of people worldwide chose whom to save across various accident scenarios. Intriguingly, the results differed somewhat by culture. Some societies tended to prioritize the young, some the many, some those who obeyed the law. This shows how difficult it is to build a "universal moral algorithm."

Moreover, real-world autonomous driving is far more complex than the textbook trolley problem. Real situations are full of uncertainty. The car cannot know perfectly whether a pedestrian is truly a person, how many there are, or exactly what will happen if it swerves. So many experts stress that "designing to prevent accidents in the first place" matters far more than "an algorithm that chooses whom to kill."

Many Perspectives, and Open Questions

Having followed this far, you may feel the answer has grown harder than when you started. That is precisely the power of a good thought experiment.

The utilitarian says, "Do not be swayed by emotion. That five lives are more precious than one is plain. Our hesitation on the footbridge may be an outdated bias left by evolution."

The deontologist replies, "Reducing everything to numbers is dangerous. If one person may always be sacrificed for the many, then none of us is safe. Human dignity holds a line that cannot be bargained away."

From the standpoint of virtue ethics, the question shifts again: "What matters is not which choice you make but what kind of person you become. In such a tragic situation, how would a compassionate and thoughtful person feel and act?"

Each of these perspectives holds a piece of the truth. Perhaps the most important lesson the trolley problem teaches us is the very fact that no single "right answer" falls neatly out. Our moral lives do not reduce to tidy formulas. They are an endless process of seeking balance among conflicting values.

Closing: Food for Thought

The trolley is still coming. And the lever is still beside your hand. Perhaps what truly matters is not whether you pull it, but what you wrestled with when you decided.

Finally, here are some questions to put to yourself.

Your intuition pulls the lever but cannot push the person. How would you justify that difference? Or is the intuition mistaken?
If the five were strangers and the one were a member of your family, would your answer change? Would that be the right thing to do?
If a self-driving car company advertised that it "always protects the occupant first," would you buy that car? What would happen to society if every company did the same?
If moral intuitions are products of evolution, should we trust them, or correct them with reason?

There are no easy answers to these questions. But simply by asking them, we become slightly deeper moral beings. Perhaps that is the real reason philosophers have crafted such riddles for more than two thousand years.