Utilitarianism: The Arguments For and Against

It has no place for the past. Promises, contracts, desert and gratitude are all backward looking, and on a strict consequentialist view they matter only through their effects. But a promise seems to bind because it was made.

It cannot accommodate special obligations. A parent's duty to their own child is stronger than to a stranger's child, and impartial aggregation cannot explain why.

Attacking hedonism

Pleasure is not the only good. Knowledge, achievement, friendship, autonomy and beauty seem valuable in themselves and not only as sources of satisfaction.

The experience machine objection. If a machine could give a person any experience they wished while they floated in a tank, hedonism says they should plug in. Almost nobody would, which suggests people value doing things and not only the experience of doing them.

Pleasures cannot be measured or compared between persons. There is no unit and no way of establishing that one person's satisfaction exceeds another's.

Attacking aggregation

This is where the decisive objections are, and an answer should say so.

It permits sacrificing an individual. If the sum favours it, a person may be used, harmed or killed for the greater good. Every version of this objection is one instance: convicting an innocent person to quell a riot; taking one person's organs to save five; permitting an industry to injure a few thousand for the benefit of a few million.

It ignores distribution. Two arrangements with the same total are equivalent on the doctrine, even if one distributes it evenly and the other concentrates misery on a minority. Aggregation is blind to who gets what.

It cannot say when to stop. There is no threshold at which a burden becomes impermissible however large the benefit. Increase the number of beneficiaries and any burden is justified.

Rawls's version, and it is the sharpest. Utilitarianism does not take seriously the distinction between persons. It extends to society a principle of choice appropriate for one individual, who may reasonably accept a present sacrifice for a greater future benefit to themselves, and applies it to different people, where the person bearing the loss is not the person receiving the gain.

Attacking impartiality

It is inhuman in its impartiality. It requires a person to be indifferent between their own child and a stranger, and treats attachment as a bias.

It is too demanding. If every rupee spent on oneself could have produced more happiness elsewhere, then almost every ordinary life is a moral failure.

It leaves no room for integrity. Bernard Williams's objection: a doctrine that requires a person to abandon their deepest commitments whenever the calculation demands it alienates them from their own projects, and a person is not merely a channel through which the best outcome flows.

The replies utilitarians make

An answer that gives the objections without the replies is half an answer.

Rule utilitarianism answers most of the aggregation objections, since rules permitting the punishment of the innocent, or the breaking of promises, would be catastrophic if generally followed.

Mill's security argument at chapter 650 does the same work within act utilitarianism: no ordinary gain can outweigh the damage to security that a practice of punishing the innocent would do.

Indirect utilitarianism. The doctrine may be true as a criterion of rightness and a bad decision procedure. Human beings calculate badly, so they should follow rules, cultivate dispositions and keep promises, because doing so produces better results than calculating case by case.

Diminishing marginal utility supplies a distributive tendency: a rupee is worth more to somebody with few, so on the doctrine's own terms an equal distribution usually produces more total happiness than an unequal one.

And a concession worth making. The objections about measurement are the weakest, because every alternative theory also requires judgment, and none of them has a procedure either.

The assessment an answer should reach

Two propositions, and both are defensible together.

Utilitarianism is indispensable as a component. Any theory that ignored consequences entirely would be monstrous in a different way, and no government can decide without comparing effects. It is what stops rights talk becoming an excuse for not thinking about outcomes.

It is insufficient as the whole. Because it counts only totals, it cannot supply the constraint that says a person may not be used, and that constraint has to come from somewhere else: from rights, from desert, or from the Kantian requirement that a person is never merely a means.

Which is exactly the position modern law occupies. Chapter 490's combined theory of punishment is utilitarianism operating within limits set by desert, and article 19's structure at chapter 620 is a public interest justification operating within a requirement of reasonableness. The mature position is not a choice between the doctrines but a division of labour.

A worked example

A vaccine will prevent a disease in ten crore people. It is known that a very small number, perhaps one in a lakh, will suffer a serious adverse reaction, and there is no way to identify them in advance.

The utilitarian analysis is straightforward and correct. The programme prevents an enormous quantity of suffering and causes a small one. It should proceed.

Notice what the doctrine got right. A rights-based analysis that forbade any imposition of risk on an unconsenting person would forbid the programme and let the disease continue, which nobody thinks is the right answer. This is the case that shows why utilitarianism cannot be dispensed with.

Now change one fact. Suppose the thousand people who will be harmed could be identified in advance, and the programme proceeded anyway without telling them.

The arithmetic is identical, and almost everybody's judgment changes. What has changed is not the total but that identified persons are now being knowingly used for others' benefit.

And what the law actually does with the original case. It proceeds with the programme, requires informed consent, and establishes a compensation scheme for those injured. That is aggregation supplying the decision and a non-aggregative principle supplying the constraint and the remedy, which is the assessment above in operation.

Distinctions that carry marks

What these arguments are NOT

They are not arguments for ignoring consequences. The vaccine example shows what that would cost.

The measurement objection is not the strongest. Students reach for it first and it is the weakest, because every rival theory requires judgment too.

Rule utilitarianism is not a complete escape. It faces the dilemma at chapter 630: either it collapses into act utilitarianism or it becomes rule worship.

"Utilitarianism justifies anything" is false. It justifies anything whose consequences are best, which is a real and often demanding constraint. The objection is that the constraint is the wrong kind, not that there is none.

Quick revision

1. In favour: a single clear standard; impartiality, which made it a reforming doctrine; secular and public reasons; consequences taken seriously; demanding rather than permissive; the working method of public policy; and progressive by construction.
2. Against consequentialism: consequences cannot be known; no place for promises, contracts, desert or gratitude; no room for special obligations.
3. Against hedonism: pleasure is not the only good; the experience machine; pleasures cannot be measured or compared between persons.
4. Against aggregation, and these are decisive: it permits sacrificing an individual; it ignores distribution; it cannot say when to stop; and Rawls, that it does not take seriously the distinction between persons, extending to society a principle of choice fit for one individual.
5. Against impartiality: inhuman between one's own child and a stranger; too demanding; and Williams's integrity objection.
6. Replies: rule utilitarianism, Mill's security argument, indirect utilitarianism, and diminishing marginal utility.
7. The assessment: indispensable as a component and insufficient as the whole, which is exactly the position of the combined theory of punishment and of article 19's reasonableness requirement.

Test yourself

1. Set out the strongest arguments in favour of utilitarianism. That it supplies a single, clear and usable standard applicable to any question, which no rival theory does. That it is impartial, counting everybody for one and nobody for more than one, which made it an instrument of reform against rank, caste, sex and property. That its reasons are secular and public and can be stated to anybody. That it takes consequences seriously, so that insisting on a rule which produces avoidable suffering must be justified. That it is demanding rather than permissive, since one's own happiness counts for exactly one. And that it is the working method of public policy, since governments must compare effects.

2. Which objections are decisive, and why? Those attacking aggregation. Because only the total is counted, an individual may be sacrificed whenever the sum favours it, which generates the objections about convicting the innocent and about imposing severe burdens on a few for a moderate benefit to many. Because gains and losses offset, distribution is invisible, so two arrangements with the same total are equivalent however unevenly one is distributed. And because there is no threshold, no burden is impermissible if the number of beneficiaries is large enough.

3. State Rawls's objection. That utilitarianism does not take seriously the distinction between persons. It takes a principle of choice which is entirely reasonable for a single individual, who may accept a present sacrifice in order to secure a greater benefit to themselves later, and extends it to a society, where the person bearing the loss is not the person who receives the gain. What makes the trade-off acceptable within one life is that the same person experiences both sides of it, and that feature is exactly what is absent when the doctrine is applied across different people.

4. What replies do utilitarians make to the aggregation objections? Rule utilitarianism, since a rule permitting the punishment of the innocent or the breaking of promises would be catastrophic if generally followed. Mill's security argument, that no ordinary gain can outweigh the damage done to the vital interest in security by a practice of that kind. Indirect utilitarianism, that the doctrine may be the correct criterion of rightness and a bad decision procedure, so that people should follow rules and keep promises because doing so produces better results than case by case calculation. And diminishing marginal utility, which gives the doctrine a distributive tendency without a distributive principle.

5. What is the fair assessment of utilitarianism? That it is indispensable as a component of political and legal reasoning and insufficient as the whole of it. Indispensable, because a theory ignoring consequences would be monstrous in its own way and because no government can decide without comparing effects, as the vaccine example shows. Insufficient, because counting only totals cannot generate the constraint that a person may not be used, which must come from rights, desert or the requirement that a person never be treated merely as a means. Modern law occupies exactly that position, with utilitarian aims operating inside non-utilitarian limits.