Utility

In plain terms

If what matters is that things go well for people, then the right thing to do is whatever produces the most of that. Count everyone, count equally, and act on the total.

The radicalism of it is easy to miss now. When Bentham wrote, counting everyone equally meant counting the poor, the imprisoned and the colonised - and he extended it to animals on the grounds that the question is not whether they can reason but whether they can suffer.

The problems are in the counting. Benefit to whom, measured how, and can one person's gain genuinely offset another's loss? A total can go up while a particular person is made much worse off, and any theory working with totals has to say why that is acceptable - which is where most of the argument has been for two centuries.

As a philosopher would put it

Bentham's felicific calculus proposes to measure pleasure and pain by intensity, duration, certainty, propinquity, fecundity, purity and extent, and to sum across all affected. Mill's revision introduces qualitative distinctions between higher and lower pleasures, on the ground that a dissatisfied Socrates is preferable to a satisfied fool - which many have argued is inconsistent with the hedonism it is meant to defend.

Modern welfare economics reconstructs utility as preference satisfaction, measurable ordinally through choice, which avoids the metaphysics of pleasure at the cost of the very interpersonal comparisons the moral theory requires. Arrow's impossibility theorem shows that no aggregation rule satisfying a small set of reasonable conditions exists for ordinal preferences.

The standard objections are aggregative. Utility monsters, repugnant conclusions and the acceptability of severe losses to a minority for small gains to a majority are all consequences of adding welfare across persons. Rawls's response is that the separateness of persons makes such trade-offs illegitimate between people even though they are legitimate within a life.

What is actually disputed

Whether welfare can be added across people.

The theory requires interpersonal comparison - that one person's loss can be outweighed by others' gains - and there is no agreed method for making it. Economics largely abandoned the attempt and works with ordinal preferences, which cannot be summed; Arrow's theorem shows the difficulty is structural rather than technical.

Even granting comparability, the aggregative structure produces conclusions its defenders accept reluctantly. Parfit's repugnant conclusion - that a vast population with lives barely worth living can outrank a smaller flourishing one - follows from plausible premises and has resisted forty years of attempted escapes.

The defence is that every alternative faces its own catalogue, and that a theory which counts everyone equally and asks what actually happens to people has a claim on our attention that no amount of counterexample fully dissolves.