The trolley problem isolates two rival decision rules for the same arithmetic of harm. Both scenarios below share one baseline: an out-of-control trolley is about to kill several people unless you intervene.
Utilitarian rule (consequentialist):
choose action a that maximises U(a) = −deaths(a)
i.e. minimise total deaths, full stop — numbers are all that matter
Deontological rule (doctrine of double effect):
action permissible only if the harm is a foreseen SIDE EFFECT,
never the intended MEANS to the good outcome
→ switch: diverting a trolley is permissible even though a death is foreseen
→ footbridge: using a person's body to physically stop the trolley
treats them as a mere instrument — impermissible regardless of the count
- Switch scenario — pulling the lever diverts the trolley onto track B; the people on B die only as a side effect of saving those on A, so most deontological theories (and most people) accept it once B's count is lower.
- Footbridge scenario — the only way to stop the trolley is to push a bystander onto the track, using their body as the stopping mass. The arithmetic can be identical to the switch case, yet intuitions — and Kantian ethics — usually reject it, because the death is now the means, not a side effect.
- Framework match readout compares the action you actually took to what each selected framework recommends for the current numbers, so you can see exactly where they diverge.
This asymmetry (same math, different verdict) is the core empirical finding behind the trolley problem since Philippa Foot introduced it in 1967 and Judith Jarvis Thomson added the footbridge variant in 1985 — it remains the standard probe for whether a moral judgment tracks consequences or the structure of the act itself.