Cui Bono?, or, Avoiding the Repugnant Conclusion

Derek Parfit has an ingenious argument[1] intended to demonstrate that our intuitions about value are badly broken. It runs as follows. Assume for convenience that we can measure the quality of everyone’s life, and that quality of people’s lives is all that matters. (We don’t need to assume that overall utility is obtained by summing individual qualities of life, or anything like that.) Now, consider a world containing N people with a quality of life Q. Clearly, says Parfit, this is no better than a world containing those people, plus M (>>N) people with quality of life Q’, slightly worse than Q but still positive (i.e., worth living). And this, in turn, is no better than a world containing (N+M) people whose quality of life is Q’’ (very slightly worse than Q but better than Q’). And this, in turn, is strictly worse than one with (N+M) people whose quality of life is Q’’’ (better than Q’’ but still a tiny bit worse than Q). In summary, a world containing "N1 at Q1" is worse than some world containing N2 at Q2 where Q2<Q1, provided N2 is very large. Now we can repeat this construction many times, eventually obtaining a world in which there are a vast number of people, all of whose lives are just barely worth living. This world, if our moral intuitions are right, can be shown to be better than the present world. This is what Parfit calls the Repugnant Conclusion; and it is indeed repugnant to most of us.

One way to avoid the Repugnant Conclusion would be to reject the technical argument I have just sketched; to deny, in other words, that we can get from the present world to the Repugnant one by a sequence of value-improving changes. However, Parfit’s presentation of the argument is (to me) very convincing, though doubtless my own abbreviated presentation above is less so. I wish to suggest another way in which we can deal with the problem: I think that if the Repugnant Conclusion is properly understood, it can be seen to be much less repugnant than at first it appears.

By way of motivation, let us digress briefly onto the question of what makes an individual’s life good or bad. Everyone agrees that, whether or not they tell the whole story, hedonistic theories provide an important part of the answer. There are many ways of stating hedonistic principles; the one I want to consider is instructively wrong.

Specifically, consider the following principle. "A measure of the value of someone’s life is given by comparing the number and strength of their satisfied desires with the number and strength of their unsatisfied desires." Superficially, this seems plausible. But here is an example that shows it to be inadequate (also from [1]). I have the option of becoming addicted to a certain drug. If I become addicted to the drug, I will have for the rest of my life a very strong desire to take more of it. (It will not make me any happier; I will just desire it for no reason other than my addiction.) Even if I am certain of always being able to afford a supply of the drug, my life would not be improved by becoming addicted to it; and yet doing so might maximise my "balance of desires" by giving me a hugely strong fulfilled desire.

What’s wrong with this theory? The answer seems to be that once a desire is there it’s good for it to be fulfilled, but that desires aren’t in themselves good. (Perhaps they are even bad in themselves, as the Stoics believed.)

None of that is either new or surprising. Let us now consider a parallel principle that is widely believed. "A measure of the value of the world is given by comparing the number and value of worthwhile lives with the number and (negative) value of not-worthwhile lives." (This is one natural interpretation of "the greatest good of the greatest number".) Perhaps this principle is flawed in much the same way as the naive version of hedonism discussed above: perhaps lives, like desires, are important once they’re there rather than good in themselves.

More specifically, consider a theory of the following kind. No intrinsic value attaches to lives as such. However, a life (even a rather wretched one) certainly has value to its bearer, which is why killing is usually wrong. (I note in passing that this probably implies some theory like that adumbrated in [2], in which one can wrong someone who has ceased to exist.)

Thus, we are distinguishing between two kinds of value a life can have: a subjective value and an objective one. Objectively, the mere fact that someone is alive is of very little value. Subjectively, it matters hugely; not because life is a great good, exactly, but because being deprived of life is a great evil. But it is perfectly possible that the following might both be true of some person: (1) It would have been better had he never existed, but (2) It would be very wrong to cause him to cease to exist.

Clearly, it is the objective kind of value that matters for Parfit’s argument: the world of the Repugnant Conclusion is one in which each life is only just worthwhile objectively: it is only just better that each person should exist than that they should never have existed. But the Repugnance of the Conclusion is the result of our assumption that a life that’s only just worthwhile would be, in fact, rather nasty. We think of a life that is only just worthwhile to its owner; a life that is only just preferable to death, only just good enough to stave off suicide. This is a mistake: a life of that quality is not objectively worthwhile at all. If given the choice of creating, or not creating, a person who will have difficulty deciding whether or not it’s worth going on living, most of us would unhesitatingly go for not creating.

So, what would the world of the Repugnant Conclusion really be like? It would have a vast population. Everyone in it would have a life objectively only just worth living: in other words, if asked the question "Would it be better to add one more person to the world if their life will be of exactly this quality?", the answer would be "yes, but only just". It seems to me that this implies quite a respectable quality of life: certainly good enough that it’s not at all obvious that such a world would be worse than the present one. I would suggest, in fact, that it’s not even obvious that most lives in the present world are objectively worth while.

One final remark. In all the above I have assumed that the existence or otherwise of a particular life is relevant only to its owner. In fact, of course, there are usually many other people who benefit from someone’s life. I don’t think this has any material effect on either Parfit’s argument or mine.

[1]Derek Parfit, Reasons and Persons.

[2]Thomas Nagel, Death (in the book Mortal Questions).