The trolley problem.

Imagine choosing between a small benefit for many residents and a major benefit for a few. Count the size of each benefit, not just the number of recipients. A majority gaining slightly can still be outweighed by severe losses imposed on a minority.

Aggregation, demandingness, and the treatment of rights are persistent challenges. Supporters can consider indirect effects, uncertainty, and the value of dependable rules. Critics ask whether even a fully informed welfare calculation could justify treating one person unfairly. The debate concerns the standard of rightness as well as usable ways of deciding.

Imagine a town deciding between accessible transport and a decorative public project. An outcome-based comparison should include who can use each project, the seriousness of the benefits, and what the town would forgo. Counting participants alone may miss a large improvement for a smaller group. The example exposes a recurring question: can a welfare calculation adequately protect people with unusual needs, or must additional principles constrain what institutions may trade away?

Imagine repairing a dangerous bridge while knowing that the noise will disturb nearby residents. Compare deliberately creating that disturbance to drive residents away. The same inconvenience can occupy different roles in a plan; proportionality and less harmful alternatives still matter.

Critics ask whether intention can bear so much moral weight or be reliably distinguished from foresight. An agent may describe a plan strategically to avoid acknowledging intended harm. Applying the doctrine therefore requires examining how the plan works, what the person is committed to achieving, and whether the harm is necessary.

Use an invented infrastructure case: repairing a dangerous bridge temporarily disrupts businesses, while deliberately closing a business to pressure its owner uses that disruption as part of the plan. Similar losses could result, but the losses play different roles in the explanation of action. In either case, ask about alternatives and avoidable burdens. Even a clearly unintended cost needs justification; classifying it correctly does not finish the assessment of what the decision maker owes.

Imagine a researcher tempted to use private records without permission because the analysis could help many people. A duty-based assessment asks what participants are owed and whether the use respects their authority, alongside any assessment of the project’s expected benefits.

Constraints create hard cases when every option infringes a duty. Some theories permit exceptions beyond a threshold; others distinguish doing harm from allowing it. A key objection asks why one should respect a constraint when violating it would prevent more violations. Different deontologists offer different answers rather than a single universal formula.

Consider a fictional official who can improve service statistics by concealing unfavorable information from applicants. A duty-based objection asks what the applicants are owed as people entitled to make informed choices. It cannot be answered solely by showing that the average outcome improved. But the objection still needs precision: which information is relevant, who has a claim to it, and whether protecting someone else’s privacy creates a competing responsibility?