Utilitarianism, Supererogation and Future Generations

I shall argue here that the reason supererogatory acts are not obligatory is that they require too much personal sacrifice [I am using "sacrifice" in a broader way than usual. I shall say that someone is making a sacrifice whenever he knowingly acts for the sake of others in away that entails personal loss for him], and that in order for an act to be supererogatory, it must have a kind of result that you would have an obligation to bring about if you could do so with no personal sacrifice. I further argue that traditional utilitarianism should be modified so as not to treat supererogatory acts as obligatory.

The first two points bear on William Anglin's criticism ["The Repugnant Conclusion", Canadian Journal of Philosophy 7 (1977), pp. 745-54] of an earlier paper of mine ["Utilitarianism: The Classical Principle and the Average Principle", Canadian Journal of Philosophy 5 (1975), pp. 409-19] where I gave the following two cases to support the contention that it is prima facie wrong to prevent the existence of future generations because it is prima facie wrong to prevent the existence of happy people.

Case 1. A woman knows that if she has a baby, he will be so seriously malformed that it is virtually certain that he will wish he had never been born. Still, she wants a baby and she cannot adopt one because she has severe emotional problems. She believes (correctly) that she would be happier with a malformed child than with no child at all. Her added happiness, however, would be small in comparison with the child's wretchedness. The disadvantage to society of her having a child who might become a public burden would be balanced by the advantage that she would become less problematic herself. In such a case, most people would, I think, agree that she should not have the baby.

Case 2. A population planner can get a birth-control program accepted in a world in which there is no threat whatever of over-population. The existent population would be about as well off with the program as without it, but although the overwhelming majority of those persons whose lives would be prevented if the plan were put into effect would have happy lives, statistical considerations make it virtually certain that some of them would have wretched lives. If advocating the program would only prevent those lives, he surely ought to advocate it, but in this case it would clearly not be wrong not to advocate the program, even though it means letting wretched people be born (in contrast to case 1 where it was clearly wrong to let the wretched baby be born). Why isn't it wrong in this case? Surely, because by advocating the program the population planner would be preventing the existence of a great many happy lives as well as a few wretched ones. It seems to follow that preventing the birth of happy people is not morally neutral but is prima facie wrong: otherwise, there would be nothing in this case to counterbalance the prima facie wrongness of failing to prevent the births of the wretched people.

I also claimed that these cases show that there is a prima facie obligation (which is obviously commonly overridden) to bring about the existence of happy people. Anglin objects that our response to such cases can be adequately explained by its being supererogatory, but in no way obligatory, to bring about the births of happy people [Jonathan Bennett raises other objections to my argument ("On Maximizing Happiness", in a forthcoming collection of papers, Obligations to Future Generations, ed. R. I. Sikora and Brian Barry (Philosophical Monographs, Spring, 1978)), to which I respond in my contribution to the same collection]. But in order for this to be so, he must find a case where if you bring about a certain state of affairs, it is a supererogatory good deed but where, if that state of affairs could have been arranged at no cost whatever to yourself, you would have had no obligation whatever to do so. I doubt that such a case can found. For instance, Schweitzer's work in Africa is a paradigm case of supererogatory behavior, but although Schweitzer may not have had an obligation to make sacrifices to help the Africans, if no sacrifice whatever had been required of him, it would have been wrong for him not to have helped them: if you can prevent or alleviate suffering at no cost to yourself, you should certainly do so. Examples can, however, be found of actions that many philosophers regard as supererogatory even though they don't believe that they have a kind of outcome for others that an agent ought to bring about whenever it costs him nothing to do so. Take the following case. Smith could make his associates feel better by smiling when he comes to work, and smiling would involve no sacrifice whatever for him. Even so, some philosophers would regard his smiling as supererogatory. But a utilitarian could hardly do so: if smiling really didn't bother him at all and it would make other people even a little bit happier, then he ought to smile for utilitarianism says that there is a general obligation to maximize happiness. Failing to smile might not be seriously wrong but it would be wrong.

Also, for an action to be supererogatory it must be morally commendable. But a utilitarian can't regard an action as morally commendable unless it has (or is designed to have) a good result, and if something is a good result a utilitarian must hold that we have a prima facie obligation to bring it about. Thus if a utilitarian holds that it is sometimes a supererogatory good deed to bring about the births of happy people, he must also hold that we have a prima facie obligation to bring about such births. This point is directed not only to utilitarians but to anyone who thinks (like W.D. Ross) that there is at least a prima facie obligation to bring about good results (and, of course, not to prevent them).

Supererogation does, however, constitute a problem for utilitarians because they are forced to categorize many actions as obligatory which would strike most of us as supererogatory. Specifically, they must classify all actions in which other people gain the least bit more than the agent loses as obligatory: they must therefore regard Schweitzer's actions as obligatory rather than supererogatory. Though this is clearly counter-intuitive, I was once inclined to accept it. After all, there is no one consistent set of moral intuitions shared by all ordinary men so that any consistent ethical system must reject some commonly-held moral intuitions. I have come to believe, however, that although it is never wrong to do the optimific thing, you only have an obligation to do it when the ratio of gain to others compared to the loss to yourself is sufficiently high. This modifies strict utilitarianism but it preserves its method of dealing with moral dilemmas: in situations where all of the alternatives seem to be wrong, it is always morally acceptable to perform the optimific action. Also, if one selects a ratio between personal loss and public gain that is not extremely egoistic, it leaves one, surprisingly enough, committed to favoring pretty much the same public policies as a strict utilitarian because when you favor a public policy that involves a private sacrifice, the public gain will almost invariably be immensely greater than your own sacrifice. Suppose that Canada could give wheat to India, that the cost to you and to each of twenty million other Canadian would be a dollar, and that this money would do roughly twice as much good in India as it would at home. The gain/sacrifice ratio would then be roughly twenty million to one: mankind would gain roughly twenty million times as much as you would lose. With a ratio like this, you would have to be extremely selfish not to be willing to make the sacrifice.

Utilitarianism without supererogation has another advantage. One of the main problems in ethics is to find reasons that may lead people who aren't particularly benevolent to make sacrifices for others. It has become clear by now that ordinary usage on the meaning of obligation-judgments is loose enough to allow considerable leeway in defining them: we are in effect forced to offer precisive definitions. If "obligation" is defined in terms of what an ideal observer who is both cognitively ideal and completely benevolent would choose to do, we might indeed have obligations to do all that traditional utilitarianism requires, including supererogatory acts. With this definition of "obligation", however, getting someone who isn't particularly benevolent to admit that he has an obligation to do something won't help much in getting him to do it. Why should he care what a completely benevolent person, however intelligent, would choose to do if he isn't completely benevolent himself? But if "obligation" were defined instead in such a way that someone can't consistently both admit that he has an obligation to do X and deny that he himself would choose to do X if he were in an ideal cognitive state, it would be much harder to admit that you have an obligation to do a thing while denying that you have a good reason to do it. (By an ideal cognitive state, I mean one in which a person's choices are not influenced by factual ignorance, by logical errors or by failing to have a vivid idea of the effect of his action on other sentient beings. It isn't logically necessary that a vivid idea should be a sympathetic one: the connection between the two characteristics is purely contingent. Benevolence is not covertly built into the definition.)

Now it seems likely that most people in a cognitively ideal state would be prepared to make sacrifices for others if the gain to others was great enough in proportion to their own loss. It also seems likely that most people would usually regard the fact that they would only choose to do X if they were ignorant of the facts, were making a logical error or weren't vividly aware of the consequences of X, as a good reason for not doing it. If so, there is a good answer to the question. “Why be moral?" Still, most of us, even if we were in the ideal state, would probably not be willing to make sacrifices for others unless they would gain considerably more than we would lose. Given this, if “obligation" is defined as I propose, traditional utilitarianism does indeed need to cut out what I take to be supererogatory demands.

I haven't specified what particular ratio of benefit to others versus loss to oneself is required in order for there to be an obligation to make a sacrifice. Obviously, for me it is the ratio an individual would choose if he were in a cognitively ideal state. Accordingly different individuals would have widely differing obligations. However ethical judgments would still be open to criticism because one could easily be mistaken as to what one would choose in an ideal state.

Alternatively, the ratio could be set a level low enough so that it would not be too demanding for anyone in the ideal cognitive state (except for a small number of people whose extreme selfishness was firmly entrenched). With such a ratio, almost anyone who was unwilling to make the sacrifices required would be unwilling only because he was making some sort of error, or because he didn't have a vivid enough idea of the consequences of his actions. Such a low ratio has the advantages that everyone would have similar obligations and almost everyone could be given good reasons for fulfilling them. But it has the disadvantages that most people would be willing to accept a more altruistic sacrifice/gain ratio and that there would still be some extremely selfish people who would not be prepared to do what they were obligated to do, even in a cognitively ideal state. Even using such a low ratio, however, is compatible with the view that optimistic actions are never wrong and that we should almost always follow the dictates of traditional utilitarianism in the public sphere.

Would these two kinds of utilitarianism without supererogation leave us with sharply diminished obligations to future generations? As far as public policy on future generations is concerned, my earlier remarks still apply: optimific social policies would almost invariably not require too much personal sacrifice in proportion to public gain to be obligatory. But on the question of whether particular couples should have children, utilitarianism without supererogation would usually allow us to follow our own desires while traditional utilitarianism (at least as I see it) often would not.

L.W. Sumner argues (in an unpublished paper) that, on the contrary, the common belief that it is ordinarily morally permissible for people to either have children or not is compatible with traditional utilitarianism. He holds that, according to utilitarianism, people who don't want children characteristically have no obligation to have them because they can use their time in other ways that would be equally beneficial to mankind. But he forgets that frequently, and perhaps usually, such prospective parents could find not only as useful things but more useful things to do than rearing children so that though they would not have an obligation to have them according to traditional utilitarianism, they would have an obligation not to have them. Utilitarianism without supererogation can avoid this counter-intuitive result by arguing that typically, if one wants children, doing something more useful instead is not obligatory but supererogatory. This is a point in its favor. However, it would also usually allow those who want to have more children than is good for mankind to do so because usually the harm they would do to others would not be very much greater than the good they would derive for themselves. Moral intuitions about this vary a good deal. It used to be regarded as all right to have as many children as you liked provided you could support them, but now an increasing number of people consider it wrong to have a large family in an already over-populated world. Though I have considerable sympathy with this, my form of utilitarianism forces me to disagree. Still it does allow me to make what I regard as this crucial point on population policy. Two of the main reasons why so many couples continue to have large families are the high infant mortality rate in certain countries and the fact that many people need children to support them in their old age. The key obligation here is to support public policies to alter these and other conditions which make it personally advantageous to have a large family. This, like other obligations to support optimific public policies, can be derived from utilitarianism even when its supererogatory aspect is removed.

Utilitarianism, Supererogation and Future Generations
Author(s): R. I. Sikora
Source: Canadian Journal of Philosophy, Vol. 9, No. 3 (Sep., 1979), pp. 461-466
Published by: Canadian Journal of Philosophy
Stable URL: http://www.jstor.org/stable/40231108