This paper made me think what there's left for Utilitarianism when "utilitarians are worried about the impracticality objection, they should not turn to expected utility utilitarianism. That theory does not provide the basis for a cogent reply to the objection." How do utilitarians defend their moral philosophy from this one?

  • By switching from act to rule utilitarianism, see IEP. Instead of discerning the utility of particular acts, one settles on the rules designed to produce "best" outcomes "on average". – Conifold Feb 11 at 4:05
  • Anyone with experience with computing or chaotic dynamics anywhere else, or even with organizational psychology, policy or law, can see how rule utilitarianism still fails to address impracticality. The human capacity to design rulesets has never managed to make deterministic systems of laws really work. It is obvious why. And this would just be a broader attempt at the same impossible problem – jobermark Feb 11 at 19:59

From the paper

Utilitarians are attracted to the idea that an act is morally right iff it leads to the best outcome. But critics have pointed out that in many cases we cannot determine which of our alternatives in fact would lead to the best outcome. So we cant use the classic principle to determine what we should do. Its not ‘‘action-guiding’’. Some take thisto be a serious objection to utilitarianism, since they think amoral theory ought to be practical and action-guiding.

Some utilitarians respond by revising the principle by appeal to expected utility. The classic principle is replaced by a new principle that says that an act is morally right iff it maximizes expected utility. Instead of responding to the objection by replacing the classic utilitarian principle with a principle requiring the maximization of expected utility, some philosophers respond by distinguishing between a criterion for rightness and decision procedure for rightness.

In short, some people say what's moral is to try guessing the best outcome, then act on it, while the classic theory says you can only judge what's moral by seeing the effects (a posteriori).

The classic option is much more powerful, if intentions of doing good mattered more than the actual result (and we know hell is paved with good intentions), then there would be no incentive to get wiser.

Becoming wiser means you are able to better predict utility outcomes. And that becomes a duty. If only acting on guessed best utility mattered, then we could stay dumb.


Welcome, Danilo

The practicality objection

The 'practicality objection' (following Feldman) is that :

in many cases we cannot determine which of our alternatives in fact would lead to the best outcome (Feldman: 49). In other words, the objection is that in many cases, classic utilitarianism ('an act is morally right iff it leads to the best outcome' - Feldman: 49) cannot provide a decision-procedure, as John Rawls called it, for ethics. A normative ethical theory such as classic utilitarianism should be action-guiding; but classic utilitarianism isn't so, because it leaves the best outcome indeterminate. (Fred Feldman, 'Actual Utility, the Objection from Impracticality, and the Move to Expected Utility', Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition, Vol. 129, No. 1, Selected Papers from the 2004 Bellingham Conference (May, 2006), pp. 49-79: 49.)

In the following extract John Broome states and considers one objection - which isn't to say there are not others - to turning to expected utility utilitarianism to overcome the impracticality objection.

The basic point is that expected utility utilitarianism provides no more of a decision procedure than does the standard utilitarianism.

Expected utility utilitarianism does not overcome the objection

Suppose there are h people. Each has preferences among a set of alternative prospects, the same set for everyone. Each person's preferences satisfy the axioms of expected-utility theory-I shall call such preferences coherent. Expected-utility theory tells us that coherent preferences can be represented by a utility function. This function assigns a utility to each prospect in such a way that, of any two prospects, the preferred one has the higher utility. The function will also be expectational, by which I mean that the utility it assigns to a prospect whose results are uncertain is the mathematical expectation of the utility it assigns to the results. If a person's preferences are coherent, there are actually many expectational utility functions that will represent them, all positive linear transforms of each other.'

Suppose there are also social preferences among the same set of prospects. If these too are coherent, they can be represented by an expectational utility function. Once again there are actually many expectational utility functions that will represent them, all positive linear transforms of each other. (John Broome, 'Utilitarianism and Expected Utility', The Journal of Philosophy, Vol. 84, No. 8 (Aug., 1987), pp. 405-422: 405-6.)

At both the individual and the social level, there will not be a unique utility function but actually many expectational utility functions; a unique course of action, derived under the conditions Broome describes, is not to be had.

Reply to save expected utility

But isn't the right response that we can be indifferent between these utility functions since all offer equally good prospects ?

Counter-reply to sink it

The conditions which expected utility utilitarianism assumes are in fact unrealistic, and therefore the theory is impractical to rely on. What are the chances that 'Each person's preferences satisfy the axioms of expected-utility theory'? That they are coherent and can be represented by a utility function 'in such a way that, of any two prospects, the preferred one has the higher utility'?

Even if the assumed conditions do apply, how realistic is it that we can collate the necessary knowledge of utility functions at the time of decision-making? And more than that, do so for every decision among the indefinitely many decisions we have to make?Very unrealistic, I suggest.


This argument applies only on the basis that an ethical theory should always clearly tell us what we should do. If we accept that an ethical theory may define what's right, and that we can live according to it the best we can, the whole argument fails.

We can compare utilitarianism with other ethical theories, so see if the demand is reasonable. Do philosophers consider ethical systems in which the right thing to do in a certain situation may not be determinable on the spot? Are there ever moral dilemmas? Cases where the right thing to do depends on exact circumstances that can't be known at the time? Are there cases where the rights and wrongs have to be balanced against each other, so that things we don't have the time to consider would tip the balance one way or another?

I'm going to suggest that requiring an ethical theory to deliver an unambiguous answer to all questions when there's limited knowledge is overly demanding.


I cannot speak to what other philosophers have said on the topic, but I have simple answer.

No one is morally perfect, but that's okay.

According to utilitarianism, an action is better if it produces more utility, and best if it produces maximal utility. So a good utilitarian should strive to do the action that has the most utility, even though they will almost always fail. Realistically, this cannot be achieved, hence the impracticality objection, but it's not best or bust. More utility is better than less, even if it is worse than maximal.

If the doing the utilitarian calculations to determine the best action for each small scenario is too involved of a process to do in a reasonable time, then utility will be maximized by following simple rules that usually produce high utility, though perhaps not maximal.

For an over simplified example, if giving full thought to a scenario yields an average of 10 utils but I can only do this 5 times a day, but doing something that seems intuitively right or follows simple rules yields 1 util on average, but I can do it 100 times a day, then I get twice as much utility from the second case.

This has the implication of something that everyone realizes intuitively: you should spend more time thinking about more important decisions and less about unimportant decisions.

This sounds like I am promoting expected utility utilitarianism, but I am really not. The best action is the one that produces the most utility, but the best way to make decisions is to determine what has the best expected outcome. In practice, there is little difference between the two since no where in utilitarianism does being "worse" have direct consequences.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.