The Trolley Problem Meets the Hospital Generator
A hospital's emergency plan has one backup generator serving two wings. During a prolonged outage the generator can power only one wing. Wing A holds 6 patients, all on ventilators, with an estimated survival probability of 0.9 if powered. Wing B holds 2 patients, both on dialysis, with an estimated survival probability of 0.95 if powered. If a wing loses power, expected survival drops to 0.2 for its patients. The plan's author, a utilitarian, proposes powering Wing A because 6 × 0.9 = 5.4 expected survivors exceeds 2 × 0.95 = 1.9. A nurse objects that the dialysis patients will certainly die without power, whereas the ventilator patients might be manually bagged by staff, and that the plan ignores this. A deontologist on the committee says the calculation is morally irrelevant because it treats patients as numbers.
1. Reconstruct the utilitarian proposal precisely and identify one empirical assumption on which its conclusion depends.
Read the worked answer
The proposal: choose the wing that maximises expected survivors. Wing A: 6 patients × 0.9 = 5.4 expected survivors if powered; if unpowered, 6 × 0.2 = 1.2. Marginal benefit of powering A = 4.2 expected survivors. Wing B: 2 × 0.95 = 1.9 if powered; if unpowered, 2 × 0.2 = 0.4. Marginal benefit of powering B = 1.5. Powering A yields the larger marginal gain, so the utilitarian selects A.
The conclusion depends on empirical assumptions, notably: (i) the survival probabilities are accurate and comparable across conditions; (ii) no alternative intervention changes the unpowered survival rate; (iii) the two groups are otherwise similar in relevant respects. The nurse's objection attacks (ii): if staff can manually bag ventilator patients, the unpowered survival rate for Wing A rises, shrinking A's marginal benefit. If manual bagging raises Wing A's unpowered survival to, say, 0.6, then A's powered benefit is 6 × 0.3 = 1.8, below B's 1.5 only marginally - and with any further improvement, B becomes preferable. The calculation is only as good as its inputs.
2. Assess the deontologist's claim that the calculation is 'morally irrelevant because it treats patients as numbers'.
Read the worked answer
The claim has force but overstates. Its force: reducing persons to expected-survivor units can license sacrificing individuals for aggregates, which Kantians reject because persons are ends in themselves, not interchangeable units. If the plan allowed harvesting one patient to save six, the objection would be decisive.
But the objection mischaracterises this case. The committee is not choosing to kill anyone; it faces a scarcity constraint where some harm is unavoidable whichever wing is powered. In such a case, consequences are not irrelevant - they are among the morally relevant considerations. A deontologist who refused all calculation would have to choose arbitrarily, which is itself a failure of concern for the patients.
The stronger deontological move is not to reject calculation but to constrain it: certain considerations (e.g., not actively ending a life, not discriminating by irrelevant characteristics such as age or disability) set side-constraints that the maximising calculation may not override. Within those constraints, expected-survivor reasoning can guide triage. So the deontologist is right that the numbers are not the whole story, and wrong that they are morally irrelevant.
3. The nurse's objection introduces a fact the utilitarian plan omitted. Explain how the committee should revise its reasoning, and what this shows about the relationship between empirical evidence and moral argument.
Read the worked answer
The committee should treat the plan's probabilities as revisable inputs, not fixed premises. The nurse's claim - ventilator patients can be manually bagged - raises Wing A's unpowered survival rate. The committee should:
- Establish the best available estimate of unpowered survival for each wing, including staff capacity and duration of outage.
- Recompute marginal benefit with those estimates.
- Check whether the ranking of wings is robust across plausible estimates; if it flips under small changes, the decision is sensitive and should be justified transparently.
- Apply side-constraints (no discrimination, no active ending of life) before maximising.
This illustrates that moral argument is not independent of empirical fact. A utilitarian conclusion is conditional on its factual premises; when the premises change, the conclusion may change. The nurse's objection is therefore not merely a factual quibble - it is a challenge to the moral conclusion. A well-run committee should document the assumptions and the sensitivity of the ranking, so that the decision is defensible even if estimates are later revised.