Principle of Procreative Beneficence

If couples (or single reproducers) have decided to have a child, and selection is possible, then they have a significant moral reason to select the child of the possible children they could have, whose life can be expected, in light of the relevant, available information to go best, or at least not worse than any of the others (p. 274).3

One implication of Procreative Beneficence (PB) is that if you are in a position to choose between having a child when you are 17 years of age or at 25, you have a significant moral reason to go for the latter option. Everything else being equal, you can expect the life of the child you bring into existence at 25 to go better than that of the child you could bring into existence at 17, because as a parent you will then be more mature and better equipped to satisfy the child's needs.4

PB is intuitively appealing. I mean, who would intuitively believe that reproducers ought to select a child, among the children they could have, whose life can be expected to be worse than those of other possible children? But it has been criticised several times over the last few years. Apart from giving a few examples, I cannot here note and comment on all the objections to PB that have been raised, but it seems to me that Savulescu and Kahane have mounted a strong defence to most of these objections. One way to indicate dissatisfaction with PB is to argue in favour of an alternative principle to PB. For instance, John A. Robertson has defended a principle to guide our reproductive choices that he calls the principle of Procreative Autonomy (PA): any procreative option selected by reproducers is morally permissible as long as it is chosen autonomously.5 However, as has been mentioned by Savulescu, one problem with PA is that it would always be morally permissible for reproducers to select the worse off child even if no one would benefit from the choice (pp. 279–80).1

Other objections criticise PB itself, without necessarily introducing an alternative principle. Michael Parker, for example, has, among several objections, argued that PB is self-defeating, because ‘… the active pursuit of the best possible … offspring, would … make it less likely that the best possible would be achieved’ (p. 282).6 It might of course be true that the active pursuit of the best possible outcome will sometimes result in a worse outcome than that achieved by not trying to pursue the best outcome. However, this may not always be the case. And why should we not try to give our children the best start in life by, say, selecting an embryo during an IVF procedure which, compared with other available embryos, does not contain any known dispositions to serious disease? Along the same lines, Savulescu argues that ‘No one suggests it is self-defeating to try to provide a good education to our children because it is self-defeating to directly aim for the good’ (p. 287).2

The following critical discussion broadly presents the second kind of objection. I do not want to defend a specific alternative principle to PB, but instead will focus on the justification of PB and some of its problematic implications. In the Why partiality? section, it is argued that the authors of PB leave us with an implausible justification for the moral partiality shown towards the child (or children) which reproducers decide to bring into existence as compared with all other individuals. This is an implausible justification because the reasons given in favour of partiality, which are based on practical reason and common-sense morality, can, at least in some situations, just as well be used to argue in favour of impartial selection. In the Elster's objection section, I want to show why Jakob Elster's recent critique of PB is in some ways problematic. More precisely, I shall argue that a counterexample designed by Elster to query PB's endorsement of partiality towards one's own children misses the target. Finally, a genuine counterexample to PB is developed and presented, and it is suggested that, because of its partiality, PB gives the wrong answer to this new counterexample.

Why partiality?

The use, made by PB, of the notion that ‘[reproducers have a] … significant moral reason to select the child of the possible children they could have, whose life can be expected ... to go best’ leaves us with at least two points that need clarification. First, what exactly do the words significant moral reason mean? And second, given that significant moral reason implies partiality towards one's own children, why this partiality?

In answering the first question it is important to notice that Savulescu and Kahane write, ‘we shall refer to them [the strength of the reasons given by PB] as significant moral reasons in order to mark their relative strength compared with competing moral reasons’. However, it is not clear what is meant by ‘significant moral reasons… compared with other reasons’. Let us first of all accept that a moral reason in favour of (or against) selection would be one based upon benefit (or harm) to individuals in terms of well-being. There might of course be other moral reasons, such as those based on concerns about autonomy or equality, but let us assume that everything else is equal when it comes to values other than individual well-being. On a first reading, there seem to be at least two interpretations of the feature that the well-being of the possible child counts for more than the well-being of others. First, the wording could allude to the idea that we have a significant moral reason to take into consideration the well-being of possible children, as they have more well-being to gain (or lose) than others in reproductive choices. However, this is an empirical claim, and it may not always be correct, as we shall see in some of the cases to come.

Second, in what is probably the fairer of the two interpretations, the notion of a significant moral reason could allude to the idea that we should accord more moral weight to gains (or losses) in the well-being of a future child C we have decided to have than we do to other persons already existing. But, again, why should C's well-being be significantly morally more important in this way? Why this moral partiality favouring C, or the relative strength of C's well-being compared to the well-being of others? Why not accept impartiality? Why is some people's (our children) well-being more important than that of others (eg, other children)?

The following example illustrates my worry and that of others (p. 403)7 with this second interpretation a little more vividly.ⁱ Imagine you and your partner have two embryos and that you have decided to implant only one of them. One has a genetic disposition for sadism; the other has a disposition for altruistic behaviour. Assume, for argument's sake, that sadists have on average better lives than altruists, but that they are often involved in violent criminal behaviour towards others, whereas altruists often help other people. PB says you have a significant moral reason to implant the embryo with a genetic disposition for sadism; but this does not seem morally right. The lesson, I think, is that Savulescu and Kahane should write merely ‘a moral reason’, not a significant moral reason. For, of course, you can have a moral reason to bring the sadist into existence if it is expected that the sadist will have a life worth living (and a better life than the other child you can bring into existence).ⁱⁱ But, consistently with this, it seems wrong to say that the parents have a significant moral reason to bring the sadist into existence—in other words, a moral reason that carries greater weight than other moral reasons.

I should caution that I am not saying that adherents of PB have to conclude that you ought to have the child with the sadistic disposition. Savulescu and Kahane both accept that there is a moral constraint that outweighs PB. The constraint is the following: ‘…we believe that PB will require most reproducers to select the most advantaged child unless doing so is predicted to lead to a very significant loss of well-being to existing people’ (p. 281).3 The reason for this constraint seems to be that it renders PB immune to the criticism that it follows, from PB, that you ought to have the sadistic child. Furthermore, PB does not oblige us to deny that other moral reasons should be taken into account. All this notwithstanding, the authors do have to say that the parents have a significant moral reason to bring the child with the sadistic disposition into existence. And again it does not seem right to claim that this reason, concerning the well-being of one's own sadistic child, is more important, or has more moral weight, than countervailing reasons connected with other people's well-being. What, then, are the arguments in favour of this partiality?

Savulescu and Kahane defend PB, and therefore the partiality of PB, by claiming that practical reasons and common-sense morality support it. According to practical reason ‘… we have reason to choose what is good, and we have more reason to prefer what is better’ (p. 281).3 So, if child A is expected to be better off (enjoy more well-being) than child B, it is irrational not to choose child A.ⁱⁱⁱ However, it is perfectly in accordance with this description of practical reason that reproducers should choose the child that can be expected to contribute more to collective well-being in the world than any other child they could have. This means that, if reproducers should follow this idea of practical reason, they will have a reason to select the child whose life is expected to go best, but not if the life of that child is expected to harm other people more than the child is expected to personally benefit. In sum, the argument from practical reason can be used to support PB and to defend principles that would guide reproducers away from PB. Practical reason could, for example, support the principle of General PB (GPB) suggested by Elster. GPB says, ‘If reproducers have decided to have a child, and selection is possible, then they have a significant moral reason to select the child, of the possible children they could have, whose life will maximise the expected overall value in the world’ (p. 483).8 GPB is a principle which, taken alone, could give the right answer in the above-mentioned case with the altruistic and the sadistic child.

The argument from common-sense morality is based on the idea that parents have reason to care about the expected well-being of their possible children, and that ‘…these reason can be seen as extensions of parent's special relations to their own children …’ (p. 283).3 However, although common-sense morality does indeed give prospective parents a significant moral reason to choose the possible child that can be expected to be better off, and in that sense displays partiality towards their own children, this may be compatible with the idea that, in some circumstances, common-sense morality requires us to sacrifice our prospective child's well-being in order to benefit others. As Elster has written, ‘Just as prospective parents hope for healthy and happy children, they also hope … that their children will not become criminals or … make the environment miserable’ (p. 485).8 This observation points in the direction that common-sense morality can support reproductive choices that run counter to PB, because it could also follow from common-sense morality that we do not always have a significant moral reason to take into account the well-being of our own possible children. Again, then, the argument proposed by the proponents of PB in defence of the principle supports, not just PB, but also principles that would guide reproducers away from PB. From a scientific viewpoint, I must acknowledge that I do not know much about common-sense morality when it comes to reproductive choices, but it seems fair to say that it would follow from common-sense morality that we do not have a significant moral reason to bring the sadist into existence.

Elster's objection

In a recent article, and very much in line with the case of the altruistic and the sadistic child, Elster has questioned PB by testing what follows from it.8 According to Elster, PB gives the wrong answer in the following donor child case:… imagine a couple who have the choice between two future children—A and B—who have equal chances for a good life, with one exception: child A's blood type makes him a universal recipient, which gives him a slight advantage in the world. The couple learns from their doctor of the existence of a severely ill child who can only be treated with stem cells from the umbilical cord of a child with a matching blood type, a blood type child B happens to have. The doctor asks the couple to choose child B, so that the ill child can be saved, but the parents, following … PB, refuse (p. 484).8

In the immediately following discussion of this case, Elster writes, ‘I take this to be intuitively wrong.’ The reason, according to Elster, for the couple to choose child A is that child A is expected to have a better life than child B, because the blood type of child A gives A an advantage if A is in need of blood transfusions. But is this a good counterexample to PB? Not really, because it does not show what it is intended to show—namely, that PB gives the wrong answer in the case in question. Recall that PB does not imply that there could be no other moral reasons at stake in reproductive choices. ^iv Compatibly with PB, there may be other such reasons, and these may outweigh the reasons mentioned in PB. Furthermore, as we have already seen, PB has the moral constraint attached to it by its authors that reproducers should not ‘select the most advantaged child if it is predicted to lead to a significant loss of well-being to existing children’ (p. 281).3 Accepting this constraint, the proponents of PB need not accept that the parents should choose to have child A. For, if the parents choose child A, it can be predicted that another child will die, as compared with the situation where child B is chosen. It is taken for granted here that this is a significant loss. If a child in need of stem cell donation is going to die because the parents have decided to have child A, the parents’ decision partly explains why a child will die—and this, remember, is a child whose life could have been saved.

Elster's case must therefore be adjusted if it is to be a plausible counterexample to the combination of PB and the moral constraint mentioned above. Imagine the following case which, in its opening details, reminds us of Elster's example:A couple has a choice between two future children—A and B—who have equal chances for a good life, with one exception: child A's blood type makes him a universal recipient, which gives him a slight advantage in the world. However, if they choose child B their doctor knows of five older children who would all be better off, as they can use some stem cells from B (but not from A) that will give each of them an advantage in the world which, all things considered, is slightly greater than the advantage A would gain. So, all things considered, a better outcome will be secured if the couple choose to have child B.

Confronted with this case, PB seems to give the wrong answer. It says the couple should choose to have child A, as they have a significant moral reason to select the child that is predicted to have the best life. And the couple would have to do this even if the benefits for child A are only slightly better than those anticipated for child B, and even if five other children, all things considered, would benefit more if child B was brought into existence. Note that the moral constraint discussed above cannot be used in this case, because the selection of child A would not lead to a significant loss in well-being for the five older children.

So PB (even when supplemented with the constraint) seems to give the wrong answer in this case. If we follow Savulescu and Kahane's justification for PB—a justification which, as we have seen, rests on practical reason and common-sense morality—we will once again see that this justification can be used to support outcomes (eg, that the reproducers in the case ought to select in favour of child B) different from the one that follows from PB. For, according to practical reason, we ought to do what is better, and this means that we ought to select child B. I do not know what would follow from common-sense morality in the revised case, but it certainly is not obvious that common-sense morality would require the reproducers to select child A. Why not benefit five children, instead of one, when the amount of well-being, all things considered, will be greater if one does this? If it were to follow from common-sense morality that reproducers ought to select child A, then we would have a conflict between practical reason and that morality. But if we must choose between practical reason and common-sense, it seems fair to say that we, as scientists or moral philosophers, should prefer rationality. A number of research studies in moral psychology and the history of moral decision-making have shown that common-sense morality is readily distorted by bias, lack of information and/or distraction by irrelevant circumstances.9 ,10