Is Utilitarianism a Convincing Moral Theory?
Utilitarianism is not a convincing moral theory. The calculation problem is largely solved by rule utilitarianism, and hedonism is rescued by preference utilitarianism — but every form of the theory shares the consequentialist criterion of rightness, which gets the direction of moral evaluation backwards. A theory that cannot distinguish an agent who tries and fails from one who does not try at all has replaced moral evaluation with outcome evaluation, and these are not the same thing.
- Utilitarianism is a normative ethical theory holding that the right action is that which maximises utility — understood as the overall good produced by an action's consequences - It is consequentialist: the rightness of an action depends entirely on its consequences, not on the nature of the action itself or the intentions of the agent - It is hedonistic in its classical forms: Bentham's act utilitarianism holds that utility is happiness, understood as pleasure and the absence of pain; Mill's qualitative utilitarianism refines this by distinguishing higher intellectual pleasures from lower bodily ones; preference utilitarianism, advanced by Singer, replaces happiness with preference-satisfaction as the relevant good - The theory also comes in act and rule forms: act utilitarianism demands we apply the utility principle to individual actions; rule utilitarianism demands we follow rules whose general acceptance maximises utility - I will consider three objections in order of cruciality: that utilitarianism faces irresolvable problems of calculation; that Mill's proof of the greatest happiness principle fails and hedonism with it; and that consequentialism as a criterion of rightness is fundamentally mistaken — evaluating states of affairs rather than actions, and judging agents by outcomes outside their control. The first is a practical problem that rule utilitarianism largely solves; the second attacks the hedonistic foundation, motivating a move to preference utilitarianism; the third attacks what every form of utilitarianism shares and cannot be escaped. I will argue that utilitarianism is not a convincing moral theory
Section 1: Problems of Calculation
- Rule utilitarianism largely solves the calculation problem — the move to societal-level rule-following is a genuine and effective response. The strong/weak dilemma is a real residual problem but not fatal: the strong rule utilitarian can reply that what makes a rule authoritative is its consequences, even if individual instances of rule-following are not evaluated consequence by consequence. The theory survives in weakened form - The calculation objection is a practical problem rather than a fundamental one — it questions whether utilitarianism can function as a decision procedure, not whether its criterion of rightness is correct. A theory’s decision procedure and its criterion of rightness are distinct: even if we cannot always calculate which action maximises utility, it might still be true that the right action is the one that does
Bentham: The Hedonic Calculus is Unworkable
- Bentham’s act utilitarianism requires us to calculate, for each action, how much happiness it will produce across all those affected, using his hedonic calculus — seven criteria including intensity, duration, certainty, extent and fecundity - This calculation is unworkable in practice for several reasons: we cannot reliably predict the consequences of our actions, especially long-term ones; subjective mental states like pleasure and pain are difficult to measure; moral situations are often time-sensitive, leaving insufficient time for calculation; and it is unclear which beings to include — if non-human animals can suffer, they must be counted, vastly complicating the calculus - More fundamentally, the hedonic calculus assumes pleasures can be quantified and compared — but it is not obvious that the pleasure of friendship and the pleasure of eating can be placed on the same scale
Mill: Rule Utilitarianism
- Mill accepted that the hedonic calculus was too impractical for everyday moral decision-making - Rule utilitarianism addresses this by shifting calculation to the societal level: rather than calculating the utility of each individual action, we follow rules which society has determined through collective experience to maximise happiness — ‘tell the truth’, ‘do not steal’, ‘keep your promises’ - Individual moral agents need only know and follow these rules; the difficult calculations have already been made through accumulated historical experience
: The Strong/Weak Rule Utilitarianism Dilemma
- Rule utilitarianism faces a dilemma between two versions, each of which has a serious problem - Strong rule utilitarianism holds that rules should never be broken, even when breaking them would maximise happiness in a specific case — but this looks more deontological than consequentialist; if we follow a rule even when breaking it would produce better consequences, we have abandoned the consequentialist commitment to maximising utility - Weak rule utilitarianism holds that rules can be broken when doing so would maximise happiness — but this collapses back into act utilitarianism, since in every case we are simply asking what maximises utility and the rules become redundant guidelines rather than authoritative standards - The rule utilitarian has no stable middle ground: either the rules are inviolable and the theory is no longer genuinely consequentialist, or they are breakable and the theory is no different from act utilitarianism
Section 2: Mill’s Proof and Nozick’s Experience Machine
- Nozick’s experience machine has genuine force — the intuition not to plug in is near-universal and not easily explained away. The hedonist might reply that people only refuse because of irrational attachment to ‘reality’ — but this is weak. It dismisses as irrational a considered, reflective preference that most people maintain even after careful thought. P4 fails - Preference utilitarianism is a genuine and effective response to Nozick. It does not merely patch hedonism but reconceives the theory’s foundational value in a way that is independently motivated — preference-satisfaction better captures the range of things people actually care about. However it faces its own problems: calculating and comparing the strength of different preferences is no easier than calculating happiness, and preference satisfaction can include preferences that seem morally perverse — Singer himself acknowledges that a torturer’s preference must be weighed against the victim’s
Mill: Mill’s Proof of the Greatest Happiness Principle
- Mill’s proof is not a logical or deductive proof but an empirical one, based on observation. He argues that just as the only evidence that something is visible is that people actually see it, the only evidence that something is desirable is that people actually desire it. - He claims that everyone desires their own happiness as an end in itself, from which he concludes that individual happiness is desirable. - He then moves from individual to general happiness, arguing that because each person’s happiness is a good to that person, the general happiness is a good to the totality of people. - Therefore, according to Mill the greatest happiness principle, which aims to maximise general happiness, is a valid moral principle.
Nozick: Nozick’s Experience Machine
- P3 is the crucial premise: that everyone desires their own happiness as an end in itself. - Nozick’s experience machine thought experiment directly targets P4 - Imagine a machine that can simulate any pleasurable experience — writing a great novel, forming deep friendships, achieving significant things. Once plugged in, you would not know you were in the machine. The experiences would feel entirely real and be entirely pleasurable - If happiness were the only intrinsic good, it would be irrational not to plug in — the machine guarantees maximum pleasure - Yet most people would refuse. Surveys confirm this intuition. We want to actually write the novel, actually have the friendships, actually make the achievements — not merely have the experience of doing so - This shows that people desire things other than happiness as ends in themselves: authentic experience, genuine achievement, real connection with others. P4 is therefore false - If P4 is false, C2 does not follow — happiness is not the only intrinsic good — and the greatest happiness principle loses its justification
Singer: Preference Utilitarianism
- Preference utilitarianism, advanced by Hare and Singer, responds to Nozick by replacing happiness with preference-satisfaction as the relevant good - The right action is that which maximises the satisfaction of preferences across all those affected - This sidesteps Nozick’s objection entirely: if someone prefers not to plug into the experience machine, their preference for authentic experience is itself what should be maximised. Preference utilitarianism validates rather than overrides that preference - Preference utilitarianism is therefore a more sophisticated form of the theory that survives where hedonism fails
Section 3: Consequentialism as a Criterion: The Wrong Direction of Evaluation
- Singer’s response is the strongest available to the consequentialist, and it has genuine force — consequences do matter, and a moral theory indifferent to whether people actually suffer or flourish would be deficient. But Singer conflates what matters morally with what determines the rightness of an action. These are not the same question. Consequences can be deeply morally relevant — something every non-consequentialist theory acknowledges — without being the sole criterion of rightness. Singer needs to show not just that outcomes matter but that they are the only thing that determines rightness, and this he does not establish - The two-agent case is the decisive objection Singer cannot answer. It does not merely produce a counterintuitive result — it reveals that consequentialism has replaced moral evaluation with outcome evaluation, and these are different things. A doctor who does everything right and loses their patient has not acted wrongly. A doctor who acts negligently and saves their patient by luck has not acted rightly. Consequentialism inverts both verdicts whenever the outcomes happen to reverse, which shows it is not tracking moral rightness at all
: Consequentialism Evaluates States of Affairs, Not Actions
- Every form of utilitarianism shares a single foundational commitment: the rightness of an action is determined solely by its consequences. Sections 1 and 2 identified problems with specific versions of the theory — both were resolved by moving to a more sophisticated version while leaving this consequentialist criterion intact. This objection attacks that criterion directly - On the consequentialist view, what is evaluated first and fundamentally are states of affairs — how the world is after the action. Actions are judged only derivatively, insofar as they produce those states. This gets the direction of moral evaluation exactly backwards - Moral evaluation is evaluation of agents and their actions — of what a person does, and the will behind what they do. Consequences are not part of the action. They are what the world does after the action is complete, shaped by countless factors entirely outside the agent’s knowledge or control. To judge an action by its consequences is therefore to judge an agent by something they have no power over - The practical implication is damning. Two agents can perform identical actions with identical intentions — one produces good consequences by luck, the other produces bad consequences through no fault of their own. On the consequentialist criterion, one acted rightly and the other wrongly. But this is not a moral distinction at all — it is merely a distinction between two different states of affairs the world happened to produce. The theory has ceased to evaluate agents and is instead evaluating outcomes they had no power over
Singer: Singer’s Defence — Morality Must Be Outcome-Oriented
- Singer argues that morality must be outcome-oriented because only outcomes affect actual beings. Intentions and actions in themselves change nothing — what changes the world, and what changes people’s lives for better or worse, is what actually happens as a result of what we do. A child does not benefit from a good intention that fails to reach them; they benefit from food, medicine, and care — all consequences - Singer therefore argues that a moral theory which evaluates anything other than actual outcomes is, at bottom, self-indulgent — more concerned with the moral purity of the agent than with the actual welfare of the people morality is supposed to protect. To focus on intentions rather than consequences is to prioritise how the agent feels about themselves over whether people actually suffer or flourish - On this view, the two-agent case simply shows that luck is morally relevant. The world the first agent produced is better, and that is what matters. Moral evaluation, Singer insists, should track the world not the agent’s inner life
: Singer’s Defence Fails
- Singer’s argument proves too much and ultimately collapses into circularity. He says consequences matter because they affect actual beings — but affecting beings is itself a further consequence. The argument is: consequences matter because of their consequences. This does not establish that consequences are the correct criterion of rightness; it merely establishes that consequences have effects, which nobody denies - More fatally, Singer cannot explain the most basic and universal feature of moral judgement: that we distinguish between an agent who tries their best and fails through no fault of their own, and one who does not try at all but gets lucky. If only actual outcomes determine rightness, these two agents have done the same moral thing whenever their consequences happen to be identical. But every moral judge — and every legal system — treats them as having done something fundamentally different. The agent who tried did something right; the one who did not try did something wrong or at best morally neutral. Consequentialism has no resources to capture this distinction because it has eliminated the agent from moral evaluation entirely - Singer calls this focus on the agent self-indulgent. But this is not self-indulgence — it is the recognition that moral responsibility requires agency, and agency requires control. Holding an agent responsible for what they could not control is not moral seriousness; it is a category error. A theory that cannot distinguish moral luck from moral action has not produced a theory of morality at all
- The calculation objection is a genuine practical problem, but rule utilitarianism largely resolves it by shifting calculation to the societal level. The strong/weak dilemma is a residual problem but not fatal — the theory survives in weakened form - Mill's proof fails at P4: Nozick's experience machine shows that people desire things other than happiness as ends in themselves. Preference utilitarianism is a genuine and effective response — it reconceives the theory's foundational value in a way that survives Nozick, though it inherits versions of the calculation problem - The most fundamental objection is that consequentialism as a criterion gets the direction of moral evaluation exactly backwards. Moral evaluation is evaluation of agents and their actions; consequences are what the world produces after the action is complete, shaped by factors entirely outside the agent's control. Two agents performing identical actions with identical intentions can produce different outcomes by luck — consequentialism delivers different moral verdicts, but this is not a moral distinction at all. Singer's defence — that only outcomes affect actual beings — conflates what matters morally with what determines the rightness of an action, and cannot explain the universal moral distinction between an agent who tries and fails through bad luck and one who does not try at all - Consequences clearly matter morally — utilitarianism is right about that. But they are not the sole determinant of rightness. A theory that evaluates actions only derivatively, in terms of the states of affairs they happen to produce, has replaced moral evaluation with outcome evaluation — and these are not the same thing - Utilitarianism is therefore not a convincing moral theory