The trolley problem is a thought experiment: a runaway trolley will kill five people unless you pull a lever, diverting it onto another track where it will kill one. Do you pull the lever? Most people say yes: one death is better than five. Then the variant: you can stop the trolley by pushing a large stranger off a bridge onto the tracks. One dies, five live. Same numbers; many people say no. The puzzle is why we treat the two cases differently — and what that reveals about moral reasoning, responsibility, and the limits of simple utilitarianism.
The experiment, from Philippa Foot and later Judith Jarvis Thomson, exposes tensions between (1) consequences (minimise harm), (2) actions vs omissions (doing vs allowing), (3) using a person as a means (the stranger is used to stop the trolley), and (4) proximity and agency (pulling a lever feels different from pushing someone). There is no consensus "answer"; the value is in the pressure it puts on our principles. If you're a pure consequentialist, the two cases should be the same. If they're not the same for you, you're drawing a line that consequentialism doesn't explain.
In strategy and business, the trolley problem appears in disguised form: trade-offs where one option harms fewer people but requires a more "direct" or morally salient action. Layoffs vs slow decline; recalling a product vs risking harm; firing one person to save the team. The model doesn't tell you what to do; it forces you to see that your intuitions may be inconsistent and that "maximise good" is not the only principle people use.
Section 2
How to See It
The trolley problem appears when a decision involves trading one harm for another, when people agree on the numbers but disagree on the right action, or when "doing" vs "allowing" seems to matter morally.
Business
You're seeing The Trolley Experiment when a CEO must choose between laying off 500 people now (one sharp action) or letting the company decline and lose 2,000 jobs over time (omission). The numbers might favour the layoff; many feel the layoff is "doing harm" and the decline is "allowing" it. The trolley structure: one action, fewer casualties vs inaction, more casualties.
Technology
You're seeing The Trolley Experiment when an autonomous vehicle must choose between hitting pedestrians or swerving and risking the occupant. The trolley problem is literally in the design: who gets harmed, and does the car "do" the harm or "allow" it? Regulators and engineers are still wrestling with the framing.
Investing
You're seeing The Trolley Experiment when an investor must choose between supporting a company that does measurable harm (e.g. externalities) and divesting, knowing the next owner may not care. Divesting feels clean; staying and pushing for change may do more good. Same tension: action vs omission, means vs ends.
Markets
You're seeing The Trolley Experiment when a regulator must choose between approving a drug that helps many and harms a few (with known risk) or delaying approval and letting more people suffer the untreated condition. The lever is approval; the trade-off is one kind of harm for another.
Section 3
How to Use It
Decision filter
"When you face a trade-off that involves harm to some and benefit to others, ask: am I treating 'doing' and 'allowing' differently? Would I accept the same outcome if it required a different action? The trolley problem doesn't give the answer; it exposes inconsistent principles so you can choose deliberately."
As a founder
You will face trolley-like choices: fire one person to save the team, discontinue a product that helps some and harms others, or take a path that benefits many and hurts a few. Your intuitions may pull in different directions. Use the trolley frame to make the trade-off explicit and to see whether you're applying consistent principles or relying on moral luck.
As an investor
When a portfolio company faces a trade-off with moral dimensions (recall, layoffs, disclosure), the trolley structure may be in play. Understand how the team reasons: are they consequentialist, or do they care about action vs omission? That predicts how they'll handle future crises.
As a decision-maker
Use the trolley problem to stress-test your principles. If you would pull the lever but not push the stranger, you're not a pure consequentialist — and that's fine. The point is to know what principle you're using and to apply it consistently. When stakeholders disagree, the disagreement is often about which principle applies (consequences vs. "do no harm" vs. fairness).
Common misapplication: Treating the trolley problem as having a correct answer. It's a probe, not a puzzle to solve. Second misapplication: Dismissing it as irrelevant to real decisions. Real decisions often have the same structure: one path harms A, another harms B; the choice is unavoidable.
Netflix's "adequate performance gets a generous severance" approach is a trolley-like choice: keep someone in a role where they're a poor fit (harm to team and company) or let them go (harm to the individual). Hastings has framed the choice in consequentialist terms: the company and the majority of employees are better off when the wrong fit is corrected. The trolley structure: one action (firing) vs allowing (keeping someone in the wrong role). Not everyone agrees; the frame makes the trade-off explicit.
Buffett has written about the ethics of closing businesses (e.g. textile mills) that could no longer compete — laying off workers to preserve capital for better uses. The trolley structure: one decision harms some (workers) and benefits others (investors, future allocation). Buffett's public framing is consequentialist and candid about the trade-off; he doesn't pretend no one is harmed.
Section 6
Visual Explanation
The trolley problem: same outcome (1 vs 5 deaths), different actions (lever vs push). Many pull the lever, few push. The gap reveals that we don't only care about consequences.
Section 7
Connected Models
Reinforces
[Utilitarianism](/mental-models/utilitarianism)
Utilitarianism says maximise total welfare; in the trolley, that means minimise deaths (pull lever, push stranger). The trolley problem shows that many people don't act as pure utilitarians — they refuse the push. Utilitarianism is one principle; the trolley exposes others.
Reinforces
Consequentialism
Consequentialism: judge actions by outcomes. The trolley problem is a test: if you pull the lever but not push, you're not purely consequentialist. Something else (doing vs allowing, means vs ends) is guiding you.
Tension
[Moral Hazard](/mental-models/moral-hazard)
Moral hazard: when one party can take risks that another bears. The trolley problem has no hidden information; it's a pure trade-off. But in real analogues (e.g. who bears the cost of layoffs), moral hazard can compound the trolley structure.
Reinforces
[Trade-offs](/mental-models/trade-offs)
Trade-offs are unavoidable when resources or outcomes are scarce. The trolley problem is the starkest trade-off: one life vs five. Making trade-offs explicit is the first step; the trolley adds the question of whether the kind of action matters, not just the outcome.
Section 8
One Key Quote
"Why is it that we feel we may turn the trolley but may not push the man? ... We need an explanation of the moral difference between these two cases."
— Judith Jarvis Thomson, The Trolley Problem (1985)
Thomson didn't claim to have the final answer; she claimed we need an explanation. The lever and the push have the same body count. If we treat them differently, we're applying a principle that isn't "minimise harm." Naming that principle is the work.
Section 9
Analyst's Take
Faster Than Normal — Editorial View
Use the trolley to make principles explicit. When you face a trade-off that involves harm to some and benefit to others, don't just go with gut. Ask: would I accept this outcome if it required a different action? If the answer is no, you're not purely consequentialist — and you should know what principle you're using so you can apply it consistently.
Doing vs allowing is a real distinction for many people. Layoffs feel different from "letting the company fail." Recalling a product feels different from "not launching it." The distinction may not be logically strict, but it's psychologically and politically real. When you're the one "doing" the harm, expect more scrutiny and more emotional resistance — from others and from yourself.
Don't hide behind inaction. "If I don't act, it's not my fault" is a common dodge. The trolley problem shows that inaction can have the same outcome as action — five dead either way. Taking responsibility for the trade-off, rather than pretending you're not involved, is the mature move.
Section 10
Test Yourself
Is this mental model at work here?
Scenario 1
A CEO can lay off 10% of the workforce now (one decision) or cut costs slowly and likely lose 30% of jobs over two years. She chooses the layoff. Critics say she 'did' the harm; she says she minimised total harm.
Scenario 2
A company chooses between two suppliers. One is cheaper but has worse labour practices. The company chooses the cheaper one. No one is physically harmed in a trolley sense.
Section 11
Top Resources
Summary. The trolley problem exposes tensions between minimising harm and treating "doing" vs "allowing" differently. Use it to make your principles explicit and to understand why people disagree on hard trade-offs. It doesn't give the answer; it clarifies the question.
Greene uses neuroscience and the trolley problem to argue for a dual-process view of moral judgment: automatic (emotional) vs manual (rational). Accessible and empirically grounded.
Ross on prima facie duties and the limits of consequentialism. Useful for understanding why "minimise harm" might not be the only principle — and why the trolley intuitions might be tracking something real.
Leads-to
Veil of Ignorance
Behind the veil of ignorance, you don't know if you're the one or the five. Rawls used it for fairness. In the trolley, behind the veil you might choose to minimise total harm (pull lever, push stranger) because you're equally likely to be any victim. The veil can push toward consequentialism.
Leads-to
[Burden of Proof](/mental-models/burden-of-proof)
Burden of proof: who must justify what. In trolley-like decisions, the burden often falls on the person who would "do" the harm (pull, push) rather than "allow" it. The asymmetry in burden tracks the doing/allowing distinction.
