Utilitarian Problems

Does "the greatest good for the greatest number" resolve tough issues like abortion and animal rights?


Practical Ethics, by Peter Singer, Cambridge, England: Cambridge University Press, 1979, 233 pp., $27.50/$6.95.

According to the ethical doctrine of utilitarianism, the good is pleasure or satisfaction and the bad is pain or dissatisfaction. The greater the balance of good over bad in the world, the better the world is, morally speaking. Morality consists in making the world as much better as we can. So right action consists in producing the best overall balance of pleasure or satisfaction over pain or dissatisfaction. In short, morally we each should do what produces the largest net heap of goodness. This, of course, is the doctrine of the greatest good for the greatest number.

This apparently simple and neatly mathematical doctrine has long been on the defensive within philosophical circles. Especially within the last decade, more and more philosophers have taken to heart objections that have repeatedly been raised against utilitarianism. These objections have always focused on the fact that horrendous actions may be productive of the best overall social balance of good over bad. In such a case, utilitarianism would morally require the horrendous action.

One standard example is the framing of an innocent person in order to stop a snowballing crime wave. Another is the infliction of public torture on a few in order to satisfy the sadistic cravings of the many. A third is the forced redistribution of property from individuals who have earned it to others who would be more happy having that property. The ultimate point of all examples of this sort is that there are considerations of justice, of desert, of rights, which utilitarianism with its single-minded focus on aggregate net result must ignore.

Another major factor in the retreat of utilitarianism has been the increased interest among philosophers in specific moral problems—abortion, civil disobedience, euthanasia, paternalism, and so on. Utilitarianism has generally been thought to be especially inept at illuminating the complex face-to-face choices that such issues involve.


There are, however, philosophers and others who stand by utilitarianism. Peter Singer is among them. In particular, Singer's Practical Ethics attempts to demonstrate its fruitfulness when applied to real-world moral issues. Singer is especially concerned with life and death issues—with the killing of nonhuman animals, with abortion, with euthanasia, and with starvation. Many of his conclusions will be shocking to conventional sensibilities, and it is a testimony to the clarity of his exposition that at least some of these shocking conclusions are rendered plausible.

Is there anything Singer can say, however, in general defense of the utilitarian perspective? Why, after all, should anyone believe that his or her moral purpose is the greatest good of the greatest number? Because, says Singer, "ethical judgments must be made from a universal point of view," which means "that my own interests cannot, simply because they are my interests, count more than the interests of anyone else." To be ethical, therefore, individuals must forgo their own self-interest. The central assumption of this argument is that, in order to be partial to oneself or one's own purposes, one must presuppose that one has some special moral status. Then, since such a presupposition violates the universal character of ethics, any partiality to oneself or one's own purposes is unethical.

But the central assumption is false. A moral doctrine that allows or requires one to be partial in some way to oneself or one's own purposes need not presuppose a special moral status for oneself: one can allow that others equally may or must be partial to themselves or to their own purposes. Thus, the view that each ought to advance his own long-term interest is fully universalizable. It ascribes no special moral status to any individual. Similarly, the view that each possesses certain natural rights and may exercise these rights in any way that does not violate the similar rights of others is fully universalizable. Yet the first of these requires and the second permits individuals to be partial to their own well-being and goals.

Singer himself half-recognizes that the universal character of morality does not specifically point to utilitarian conclusions. He allows that "there are other ethical ideals—like individual rights, the sanctity of life, justice, purity and so on—which are universal in the required sense." Yet he goes on to assert that focusing on this universal character permits one to "very swiftly arrive at an initially utilitarian position." This is disingenuous. Nothing in the argument justifies arriving initially at any one of many incompatible universalizable ethical positions. Furthermore, to arrive at utilitarianism initially is to arrive at it permanently, for all proposed departures toward alternative theories will have to be rejected on the grounds that theories so distant from and incompatible with utilitarianism could not be correct.

On the utilitarian point of view, what is morally significant about persons is not their rationality, their capacity for choice, or even their capacity for moral and immoral action. Rather, what is significant is their capacity for experiencing pleasure and pain and satisfaction and dissatisfaction. These capacities, however, are quite fully shared with (at least) all higher members of the animal kingdom. So utilitarians have always been most open to considering the experience of all sentient beings as intrinsically morally significant.


In Practical Ethics Singer continues the argument, begun in his previous Animal Liberation, that our current treatment of animals, especially their inhumane production and slaughter for food, generates far more pain than pleasure and must therefore be condemned. The pleasures that some human beings derive from eating flesh cannot overbalance the pain inflicted in the process of providing this flesh.

Is there any fundamental difference between the killing of a human being and, say, a cow? There are two differences to which Singer can point, but neither seems to be fundamental. The first is that cows are generally unaware of the policy of killing cows; so at least until they enter the slaughter pens they do not live in fear and anxiety. Since rational beings such as most members of the species homo sapiens would, however, become aware of a (non-secret) policy of killing humans, such a policy would produce fear and anxiety. The second difference is that rational self-conscious beings not only prefer pleasure to pain and satisfaction to dissatisfaction; they also tend to prefer continued life to death. The satisfaction of this somewhat special preference is to be counted within Singer's utilitarian calculations. These two factors make the moral cost of killing rational self-conscious beings somewhat greater—how much greater is hard to say—than the moral cost of killing cows.

We can easily see here that neither of these special considerations against killing rational self-conscious beings applies to the killing of fetuses. Indeed, according to Singer, at relatively early pre-sentient stages, fetuses have no moral standing whatsoever. As sentience appears, fetuses acquire some moral standing. Still, "the calf, the pig and the much derided chicken come out well ahead of the fetus at any stage of pregnancy." Although for Singer the intrinsic evil of killing an advanced fetus is as great as the evil of killing a fish or perhaps even a chicken, this does not imply that Singer condemns abortion. For the benefits to be gained from the killing of unwanted or endangering fetuses are vastly greater than those gained from the killing of fish or chickens.

It will be noted that infants also are not self-conscious or rational. Therefore, they cannot have a desire for continued life and cannot be made anxious by a policy of killing infants. So, Singer concludes, the killing of an infant involves no more intrinsic evil than the killing of that much-derided chicken. Infants, of course, are often highly valued by their parents, which provides an extrinsic reason for not killing them. Usually, then, infants should not be killed for the same reason that beloved pets should not be killed. When they are not so valued but rather are the source of pain and the frustration of interests, the utilitarian calculation justifies infanticide.


Utilitarianism prescribes the creation of the largest net heap of satisfactions. But abortions prevent persons from coming into existence with lives involving a certain amount of satisfaction. Should not the amount of satisfaction thereby lost to the world count against allowing abortions? Singer's argument in favor of abortion tacitly assumes that it should not. Similarly, on Singer's view, the happiness of persons who could be created but do not yet exist does not count as a factor in favor of their creation. In short, as far as persons are concerned, Singer opts for what he calls the "prior existence" version of utilitarianism: it is the satisfaction or dissatisfaction of the interests of already existing people that counts.

This "prior existence" stance is apparently helpful to Singer in deflecting another traditional charge against utilitarianism—that it views individuals as replaceable units. For instance, it is often pointed out that if total results are what count, then if I could painlessly kill one person but also create another person whose prospects were as happy as the deceased, my act would be morally neutral, for it would have no net effect on the balance of satisfactions. So it seems that if total results count, people are replaceable. A person is just a receptacle for certain satisfactions, and any similar receptacle is just as good from the so-called moral point of view. In opting for the "prior existence" version, however, Singer can claim that my killing would be wrong because it undercuts the satisfactions of an already existing person, and the similar satisfactions introduced into the world by my creation of a replacement do not counterbalance this loss.

But how can Singer justify the "prior existence" stance? To some extent he believes that it is the desire for continued life in self-conscious beings that rules out replaceability. Yet the creation of another whose desire to continue living will be satisfied does, from the utilitarian moral perspective, completely compensate for the death of a self-conscious being. Alternatively, Singer invokes the idea that self-conscious individuals "are individuals, leading lives of their own" and that this is why they are not replaceable. This sounds far more plausible. And, fittingly, it has no place at all within any utilitarian theory. For this claim is nothing short of the insight that for each person his own life and purposes are special and that, from the special perspective of each lived existence, replacement is never sufficient compensation. So Singer's attempt to justify "prior existence" utilitarianism involves, in reality, an abandonment of utilitarianism.

In a final application of utilitarianism, Singer urges vastly greater aid, both private and governmental, to prevent starvation and malnutrition around the world. The cost to me of giving up color TV, dinners out, and so on is very much less than the benefits that could be provided to others with the money used to support my luxuries. Hence, this is part of my utilitarian duty.

Assuming that relief would have the projected effects (a large assumption), this is a thoroughly sound argument—given the utilitarian premise. Singer, however, also seeks to press this radically redistributive conclusion upon those who do not share this shaky moral premise. He does this by claiming that there is no intrinsic difference between killing and letting die. Letting another die when I could save that person is just another (often less direct and less malicious) form of killing. Thus, to fail to redistribute resources that would prevent death and suffering is to cause that death and suffering.

Singer, however, is gravely mistaken in equating the failure to aid with the infliction of harm. What happens to the distressed Indian beggar whom I do not aid is not a consequence of my omission. We can see this by noting that the unhappy outcome suffered by the beggar would have occurred even if I did not exist. So neither my existence nor any of the activities that make it up can be among the conditions bringing about the beggar's demise.

Singer has a response to this argument: "It is true that the person would have died even if I had never existed, but what is the relevance of that? The fact is that I do exist, and the consequentialist will say that our responsibilities derive from the world as it is, not as it might have been." But this response does miss the relevant point. By showing that some outcome would have occurred even without my existence, I do show that it does not depend upon me. So even though the facts are that I do exist and that I could have averted the starvation, we cannot conclude that I cause the outcome that would have occurred just the same had I never seen the light of day.

Here, as in his general arguments for utilitarianism, Singer fails to establish that each of us should forgo our own happiness when this would advance net, cross-species, global satisfaction. Utilitarianism has been put to the test of real-world problems, but it doesn't pass.

Eric Mack teaches philosophy at Tulane University. He recently spent a semester's leave of absence in Denmark working on the foundations of natural rights theory and the nature of freedom.