Preference Utilitarianism

Core Idea

Preference utilitarianism holds that the right action is the one that maximally satisfies the preferences of all affected parties, rather than maximizing pleasure or happiness as classical utilitarianism requires. R.M. Hare developed a two-level version grounded in the logic of moral language, arguing that universalizability forces us to weigh everyone's preferences equally. Peter Singer adopted preference utilitarianism to avoid the difficulties of measuring subjective pleasure and to extend moral consideration to any being with preferences, including non-human animals. The view faces challenges including adaptive preferences (people whose preferences have been shaped by oppression), the aggregation problem (how to compare strength of preferences across individuals), and whether satisfying uninformed or malicious preferences genuinely produces moral value.

How It's Best Learned

Read Hare's Moral Thinking (especially the two-level theory) alongside Singer's Practical Ethics. Then construct a case where preference utilitarianism and hedonistic utilitarianism diverge—such as a person who prefers a harder but more meaningful life over a pleasurable but shallow one—and trace how each theory evaluates the options.

Common Misconceptions

• Preference utilitarianism does not endorse satisfying any preference whatsoever; most versions exclude preferences based on misinformation or preferences that are themselves about frustrating others' preferences.
• The shift from pleasure to preferences is not merely terminological; it changes which beings count morally (anything with preferences, not just anything that feels pleasure) and how we measure value.

Explainer

You already know that classical utilitarianism identifies the right action as the one that produces the most happiness or pleasure — maximizing the balance of pleasure over pain across all affected beings. Preference utilitarianism keeps the core utilitarian structure (aggregate welfare across all affected parties, choose the maximizing action) but replaces the currency. Instead of asking "how much pleasure results?", it asks: "whose preferences are satisfied or frustrated, and to what degree?" A preference, in this context, is any pro-attitude — a desire, want, or interest that a being has about how things go.

The motivation for this shift comes partly from R.M. Hare's work on the logic of moral language. Hare argued that moral judgments are universalizable: if you judge that action X is right, you are committed to judging that any relevantly similar action in relevantly similar circumstances is right. This universalizability, he argued, forces us to give equal weight to everyone's preferences when deliberating. Peter Singer extended the view to its logical boundary: if moral consideration flows from having preferences (rather than from having the capacity for pleasure), then any being with preferences — including many non-human animals — deserves moral consideration. A mouse that prefers not to suffer has interests that count, not because it feels pain (though it does), but because frustrating its preferences is the relevant harm.

The key advantage over hedonistic utilitarianism is measurement. Pleasure is subjective and arguably impossible to compare across individuals — how do you weigh your headache against my joy? Preferences, by contrast, can sometimes be revealed through behavior and choice. If you choose X over Y when both are available, that's evidence you prefer X. Revealed preference is an empirical foothold that hedonism lacks. This connects preference utilitarianism to welfare economics, where economists measure "utility" through revealed choices rather than introspective reports of happiness.

But the view faces serious internal challenges. Consider adaptive preferences — people raised in conditions of oppression, deprivation, or constrained options may adapt their preferences to what is available, coming to prefer things they would not have preferred in better circumstances. A person conditioned to prefer subordination has that preference satisfied, but we might think satisfying it doesn't genuinely benefit them. This suggests raw preference-satisfaction isn't the right target. Most preference utilitarians respond by restricting the view to *informed* or *rational* preferences — what you would prefer if fully informed and thinking clearly. But this move reintroduces the measurement problem: we're back to needing an external standard of what counts as a well-formed preference.

The deepest tension is between the theory's clean consequentialist structure and the messiness of actual preference content. Suppose someone has a strong preference that a specific tree on the other side of the world never be cut down — they never see it, never interact with it, but the preference is intense. If that tree is cut down, is this a moral harm? Satisfying or frustrating such "non-experiential" preferences seems strange if we're ultimately motivated by welfare. This is why some philosophers argue preference utilitarianism must be restricted to preferences whose satisfaction the agent can in some sense experience. Where that line is drawn determines much about the theory's practical implications — and remains genuinely contested.