Ideas & Beliefs

What Is the Trolley Problem?

By BodaWiki EditorialPublished 4 min read

In short

The trolley problem asks why it seems acceptable to divert a runaway trolley so it kills one person instead of five, but not to deliberately harm one to save five.[1] Philippa Foot described the original case in 1967.[1] Judith Jarvis Thomson, who used the name in 1976 and 1985 articles, discussed versions such as pushing a man off a footbridge.[2][4]

An elevated view of an empty railway junction among misty green fields, where a single track splits into two tracks that curve apart.
AI-generated illustration

Key facts

First described by
Philippa Foot, 1967[1]
Thomson's articles
"Killing, Letting Die, and the Trolley Problem" (1976); "The Trolley Problem" (Yale Law Journal, 1985)[4][2]
The basic choice
Divert a runaway trolley so it kills one person instead of five[1]
Well-known variations
Bystander at the switch, transplant, footbridge ("Fat Man"), loop[2][5]
Ideas it tests
Doing versus allowing harm; the doctrine of double effect[3][4]
Moral Machine survey (2018)
Nearly 40 million decisions by over 2 million online participants, from 233 countries[8]

Where did the trolley problem come from?

The puzzle began with the philosopher Philippa Foot, in her 1967 article "The Problem of Abortion and the Doctrine of the Double Effect," published in the Oxford Review.[1][4] She described a runaway trolley bearing down on five people that could be steered onto another track, where it would kill only one.[1]

In Foot's version the trolley's driver decides, and she paired the case with a judge who must either frame and kill an innocent man or let five innocent people be killed in a riot.[3] Foot judged that killing was wrong in the judge's case but not in the driver's.[3] Her explanation was the difference between doing and allowing harm: the judge chooses between killing one and merely letting five die, while the driver chooses between killing one and killing five.[3]

How did Judith Jarvis Thomson change it?

Judith Jarvis Thomson took up the puzzle in a 1976 article, "Killing, Letting Die, and the Trolley Problem," and again in a 1985 Yale Law Journal article, "The Trolley Problem."[4] In the 1985 paper she said a number of its cases came from the 1976 one.[2] There she set Foot's driver beside a surgeon who could save five dying patients by killing a healthy young man, against his wishes, for his heart, lungs and kidneys, a contrast she called "Mrs. Foot's problem."[2] Everyone she asked said the driver may turn the trolley but the surgeon may not operate, although in both cases one person dies and five live.[2]

Thomson then considered a "Bystander at the Switch": the driver has fainted, and a passer-by can throw a switch to turn the trolley.[2] Doing nothing would only let the trolley hit the five, yet Thomson still thought he may turn it, which undercut Foot's explanation.[2][3] She asked why the bystander may turn the trolley when the surgeon may not operate, and was inclined to call this "The Trolley Problem."[2]

Another variation, which she named "Fat Man," puts you on a footbridge over the track, next to a very large man leaning over the railing.[2] A little shove would send him into the trolley's path and stop it before it reaches five workmen, but everyone Thomson asked said this was not permissible.[2]

CaseThe choiceReaction reported by philosophers
SwitchTurn the trolley onto a track with one personMost people regard it as permissible[5]
TransplantA surgeon kills one patient to save five with his organs"The universal reaction is condemnation"[5]
Footbridge ("Fat Man")Push a man into the trolley's path to stop itBy and large condemned[5]
LoopThe side track loops back, so only the one person's body stops the trolleySome feel more discomfort, but Thomson thought the agent may proceed[2][5]

In 2008 Thomson changed her mind, arguing that it would be wrong for the bystander to turn the trolley.[3][4]

What is the trolley problem meant to test?

Philosophers test rival moral principles, such as deontology and consequentialism, by checking whether they get important cases right.[7] The switch and footbridge cases have the same outcome, one death instead of five, so if people's reactions count as evidence, they seem to tell against the view that only outcomes matter.[7]

One question is whether doing harm is morally worse than merely allowing it: consequentialists say no, and their opponents almost all disagree.[3]

Another is the doctrine of double effect: the idea that it can be permissible to cause a harm as an unintended, foreseen side effect of a good action when causing the same harm as a means would not be.[4] On this view, the death on the side track is a side effect of diverting the trolley, while pushing a man uses his death as the means of saving the five.[4] Philosophers disagree on whether this really explains the cases.[4]

How do utilitarians and deontologists answer it?

For consequentialists, only consequences can justify an act, so consequentialism would permit pushing the man off the footbridge if his body would halt the trolley.[5] Act utilitarianism even implies that the surgeon should kill one healthy person to save five; rule utilitarians reply that a rule allowing this would not maximize utility, because people would stop trusting doctors.[6]

Deontologists stand opposed to consequentialists, and some explain the cases with a ban on using people: in Transplant and Fat Man the victim's body is needed to save the others, while in the switch case the five would be saved whether or not the one person were there.[5] Thomson's own account rested on rights, since shoving a person is already an infringement of his rights.[2][6]

What do experiments and surveys show?

The philosopher-neuroscientist Joshua Greene and colleagues ran an influential brain-imaging study of moral judgment, and his scenarios, derived from Foot's and Thomson's cases, became standard across much of moral psychology.[7] Greene argued that utilitarian judgments, which accept sacrificing one to save more, go with activity in cognitive brain areas, and deontological judgments with emotional areas.[7] He has since softened this view a bit, and his claim that these findings give reason to distrust deontological judgments is still hotly debated.[7]

MIT researchers turned such dilemmas into the Moral Machine, a multilingual online game about choices driverless cars might face.[8] Its results, published in Nature in October 2018, drew on nearly 40 million decisions by over 2 million online participants in 233 countries.[8] The responses were spread unevenly: only 130 countries produced 100 or more, and 491,921 respondents gave demographic data.[8] Among those who took part, the strongest preferences were to spare humans over animals, many lives over few, and the young over the old.[8] Lead author Edmond Awad said these main preferences were "to some degree universally agreed upon," but that the degree of agreement varied, including between clusters of countries the researchers called western, eastern and southern.[8]

Some philosophers urge caution: the Stanford Encyclopedia of Philosophy's entry on double effect asks whether such abstract trolley cases produce judgments reliable enough to guide the programming of self-driving cars.[4]

Frequently asked questions

Who came up with the trolley problem?

Philippa Foot described the original case, with a trolley driver, in a 1967 article on abortion and the doctrine of double effect.[1][3] Judith Jarvis Thomson used the name in the title of a 1976 article and again in a 1985 Yale Law Journal article, "The Trolley Problem," in which she discussed the bystander and footbridge versions.[2][4]

What is the footbridge version of the trolley problem?

You stand on a footbridge above the track, and the one sure way to stop the trolley before it kills five workmen is to drop something heavy in its path, such as the very large man leaning over the railing beside you.[2] Everyone Thomson asked said this would be wrong, even though the outcome, one death instead of five, is the same as in the switch case.[2][7]

What would a utilitarian do in the trolley problem?

Consequentialism, which lets only results justify an act, would permit pushing the man off the footbridge if his body would stop the trolley.[5] Rule utilitarians judge general rules rather than single acts, and argue that a rule letting doctors kill one patient to save five would not maximize utility, because people would stop trusting doctors.[6]

Is there a right answer to the trolley problem?

Philosophers still disagree.[4] Some take the common reactions as evidence for the doctrine of double effect, others reject that explanation, and Thomson herself argued in 2008, against her own earlier papers, that it would be wrong for the bystander to turn the trolley.[3][4]

What does the trolley problem have to do with self-driving cars?

A car swerving to avoid a crash might face a choice that looks ethically similar to swerving the trolley.[4] MIT's Moral Machine, an online game, gathered nearly 40 million decisions from over 2 million participants about how driverless cars should choose, though only 130 countries produced 100 or more responses.[8] Some philosophers question whether such abstract cases give reliable guidance for programming these vehicles.[4]

Sources

  1. [1]Philippa Foot · Stanford Encyclopedia of Philosophy · accessed October 5, 2026
  2. [2]The Trolley Problem (The Yale Law Journal, Vol. 94, 1985) · Judith Jarvis Thomson / Yale Law School Open Scholarship Repository · accessed October 5, 2026
  3. [3]Doing vs. Allowing Harm · Stanford Encyclopedia of Philosophy · accessed October 5, 2026
  4. [4]Doctrine of Double Effect · Stanford Encyclopedia of Philosophy · accessed October 5, 2026
  5. [5]Deontological Ethics · Stanford Encyclopedia of Philosophy · accessed October 5, 2026
  6. [6]Act and Rule Utilitarianism · Internet Encyclopedia of Philosophy · accessed October 5, 2026
  7. [7]Morality and Cognitive Science · Internet Encyclopedia of Philosophy · accessed October 5, 2026
  8. [8]How should autonomous vehicles be programmed? · MIT News, Massachusetts Institute of Technology · accessed October 5, 2026
Cite this article
APA
BodaWiki Editorial. (2026, October 6). What Is the Trolley Problem? BodaWiki. https://bodawiki.com/wiki/what-is-the-trolley-problem
MLA
“What Is the Trolley Problem?” BodaWiki, 6 Oct. 2026, bodawiki.com/wiki/what-is-the-trolley-problem.
Chicago
BodaWiki Editorial. “What Is the Trolley Problem?” BodaWiki. October 6, 2026. https://bodawiki.com/wiki/what-is-the-trolley-problem.
  • trolley problem
  • ethics
  • thought experiments
  • utilitarianism
  • deontology
  • moral psychology