Fetal Abnormality and Consistency
We have already seen how the expressivist argument has been deployed against the practice of SSA. But if the expressivist objection is persuasive, it seems to apply with equal force to selective abortion on the ground of future disability. As the philosopher Jeremy Williams comments, the expressivist objection is ‘widely known as the signature argument of the disability rights critique of selective abortion.24 He writes:
In the hands of disability theorists, the objection avers that the availability and widespread use of selective abortion for fetal impairment reflects and reinforces a number of hurtful and pernicious stereotypes about disabled people — for instance that they have no right to
- 22 Tabitha Powledge, ‘Toward a Moral Policy for Sex Choice’ in Neil Bennett (ed), Sex Selection of Children (Academic Press 1983).
- 23 ibid.
- 24 Jeremy Williams, ‘Sex-Selective Abortion: A Matter of Choice’ (2012) 31 Law and Philosophy 125, 144.
exist; that their lives are uniformly so riddled with misery as to be not worth living; that they bring only unhappiness and no rewards to those around them; and so on. The broadcasting of these ideas is seen both as harmful and disrespectful to disabled people in itself, and as ratifying existing societal prejudices against the disabled, thereby increasing their vulnerability to abuse.   
Against the expressivist objection as applied to FAA, some have attacked the notion that an individual choice by a woman not to bear a disabled child carries with it the pernicious suggestion that disabled people’s lives are inferior. In essence, such counter-arguments contend that a personal decision by a woman to spare herself the burden of raising a disabled child does not communicate any belief about the lesser value of disabled peoples’ lives. Likewise, it might be argued that willingness of the state to provide special permissions for FAA does not signal the view that the lives of disabled people are of unequal value, but aims only to protect the interests individual pregnant women may have in obtaining abortion.26
I do not find these sorts of responses to the expressivist objection wholly satisfactory^7 It goes without saying that pregnant women seeking out FAA are motivated by considerations of personal wellbeing and are unlikely to harbour the belief that disabled lives lack equal value. But the expressive significance of an act can depend upon more than the motivations and beliefs of the person carrying it out. Although the decision by a woman to abort a defective fetus does not necessarily convey the belief that disabled persons’ lives are unequal to others or not worth living, it does express the considered belief that, for her, the burdens of raising a child with the relevant condition will not be worth the gains. The fact that the decision is a personal one about her own approximation of the benefits and burdens does not preclude it from making any more general statement about the rewardingness of raising disabled children.
The second main argument against FAA is the consequentialist concern that discrimination against disabled people can be aggravated by the practice. One worry is that the routine termination of defective fetuses will diminish the quality of life of disabled persons by depleting their numbers and, as a consequence, reduce the incentive for social accommodation. With fewer disabled people in the world, the motivation to adapt our environment so as to counter the disadvantage experienced by disabled persons who do exist may be weaker. Some have termed this the ‘reduced resources argument’^8 For example, when the number of wheelchair-users is high, there is a greater incentive for the state, and others, to invest in the infrastructure that allows wheelchair access, like ramps or disabled bathrooms, but less incentive when there are fewer wheelchair-users requiring such facilities. The reduced resources argument seems to be strengthened when conjoined with a social model of disability, which views much of the disadvantage attending disability as a product of social discrimination rather than biological difference. If the disadvantages of being disabled are predominantly a function of social discrimination, then removing pressure to address that discrimination by reducing the numbers of disabled people will make a significant contribution to them.
As with the expressivist objection, the reduced resources argument has come in for some criticism by those who think it exaggerates the impact of reducing the numbers of new disabled people on the social inequality of existing ones. Some authors have even suggested that in some scenarios, reduced numbers of existing disabled people will augment the resources available to each79 Williams’s own conclusion is that even if the fears captured by the expressivist and reduced resources concerns are credible, it imposes too heavy a burden on individual pregnant women if their reproductive rights are abrogated so as to avoid them.  This calculation may be ultimately correct. However, whatever one makes of the standard objections to FAA, the interesting challenge remains for the person who invokes such arguments in support of a ban on SSA but does not extend the same reasoning to abortion for fetal abnormality.
As Williams notes, the expressivist objection seems to apply in much the same way to both forms of selective termination. If the reason for prohibiting SSA is its propensity to further entrench sexism and worsen the situation of women, we will want to know why FAA is not ruled out for the same reason: that it threatens to exacerbate discrimination against disabled people, partly by communicating the view that disabled peoples’ lives are not of equal value. Alternatively, if the concern about further entrenching discrimination is not weighty enough to justify prohibiting FAA, then perhaps it is not weighty enough to warrant the prohibition of SSA, where the burdens on the prospective parents are comparable. As was seen, some have attempted to counter the expressivist objection in the case of FAA by claiming that the personal decision not to become a parent to a disabled child does not express any general belief about the lesser value of disabled lives. But if this counterargument is valid, it should be able to shield SSA from the same objection. Recall the example situations of Women A, B, and C. Here, at least, one could certainly make the case that in choosing SSA the women are not expressing the belief that girls are inferior to boys, but only acting to preserve their own wellbeing in difficult circumstances. If the abortion request and the discriminatory statement can be separated out in the one case, why not in the other? Furthermore, what Williams terms the ‘strength in numbers objection’ to SSA—the worry that reducing the ratio of women to men will diminish women’s political power—seems to be far more powerful when applied to the case of FAA. Since there are far fewer disabled persons to begin with than there are women, it is reasonable to think that a significant reduction in their numbers through selective abortion will have a disproportionately negative impact on their social equality, if we accept the relationship between their numbers and their degree of marginalization.
The key question is therefore why the standard objections to permitting SSA do not apply with the same force to permitting FAA. Or, on the other side of the same coin, why are the answers to those objections not equally valid answers in the case of SSA? Most importantly, this includes the answer that, whatever its negative effects, withdrawing the option of selective termination unfairly burdens individual women who will suffer as a result, and who are not personally responsible for existing prejudices. Unless the two cases can be distinguished in some way, consistency requires that we adopt the same attitude towards both kinds of selection. This of course runs directly counter to popular moral thinking, which sees SSA and FAA as occupying opposite ends of the spectrum of moral defensibility.
One way of answering this consistency problem would be to concede there are no principled differences between SSA and FAA, and to suggest that in each case the correct conclusion will simply be a function of how the relevant concerns balance against women’s interests in reproductive freedom, taking into account the specific surrounding context. Where the culture is such that sexism and sex inequality are extremely thoroughgoing, the potential additive effects of SSA may be worse than in more equal societies. On the other hand, the burdens for individual women of not being able to obtain SSA are also likely to be more severe. Likewise, the situation of disabled persons in a given jurisdiction and the depth of existing discrimination and disadvantage may properly influence how one balances the negative effects of FAA against the imposition of mandated pregnancy on women. Perhaps it simply cannot be said, in the abstract, whether there is a strong case for prohibition in each instance. But if the devil is indeed in the detail, then it is thinkable that one could come to different conclusions on each case without being implicated in inconsistency. It does not seem correct however, that SSA is inherently more problematic than FAA—the standardly accepted position.
Is it possible to distinguish the cases of sex-selective and fetal abnormality abortion in a more general way? One possible argument might be that raising a child with a serious disability is far more burdensome than raising a child of an undesired sex. This, however, clearly does not hold true across all cases. Depending entirely on the particular case, the burdens of giving birth to a female child may be far more severe than, say, raising a child with Down’s Syndrome (consider again the example cases of women A, B, and C).
Another possible difference is that while the burdens of caring for a disabled child are objective or absolute, the burdens for women associated with producing a child of the ‘wrong’ sex are entirely a product of the social conditions with which they are enmeshed. If this is correct, then perhaps we would be justified to prohibit sex selection in an effort to counteract the social attitudes responsible for the burdensomeness of producing the ‘wrong’ sexed child, whereas prohibiting FAA will not go any way towards alleviating the burdens of raising disabled children. How convincing one finds this point of contrast will depend largely on one’s conception of disability. If disability is predominantly a social construction, then the inequality of disabled people will mirror the inequality of women in being socially created, as will be the burdens connected with fetal abnormality pregnancies.
In this respect, then, the comparison of SSA with FAA will turn on whether one accepts or rejects the social model of disability, and to what extent one thinks the burdens of raising disabled children are socially determined and, in theory, eliminable. The debate about the nature of disability is, of course, complex, and I have no intention here of offering a theory of how disability yields disadvantage. However, it is at least clear that the disadvantage entailed by disability is socially imposed to a significant degree, even if it is also, in part, biologically determined, and that positive social attitudes towards disability can help alleviate that disadvantage. The comparison between the burdens of giving birth to a girl in a sexist society, or sub-culture, and those of giving birth to a disabled child therefore goes quite far. And if it is legitimate to restrict sex selection choice as part of the effort to combat the devaluation of women, then it is presumably legitimate to prohibit FAA if it would help counter disability discrimination, if and where the burdens on individual women are roughly comparable. 
Next, it does not seem open to us to distinguish the two cases by saying that it is bad for a disabled child to be born but that this is not true of a healthy female child. This would be to invoke the fetal interests argument in respect of disability selective abortions but to argue that it is not applicable to sex selective ones. There are multiple problems with this move. Firstly, it can undoubtedly be far worse to be a girl or a woman in a particular context than to be a disabled person in a different one. These comparative determinations will always depend on many variables: how bad is the disability? How sexist the culture? How discriminatory towards the disabled? And so on. Moreover, even if it were generally worse to suffer from a certain disease or disability than to be a female in an oppressively sexist culture, the fetal interests argument still suffers from the fact that barely any disabled states can be said to amount to a life not worth living. As we saw, this is a condition for the application of the fetal interests argument in favour of selective abortion.
But the fetal interests argument suffers from further drawbacks in the way of the ‘non-identity problem’.32 Non-identity problems arise in ethics where it is claimed that a certain act or practice is wrong because it harms a person, but where it is difficult to identify an individual subject of harm. If fetuses are not yet persons, as we are presuming they are not, and not capable of being harmed, then they cannot be harmed by being allowed to live and become persons with difficult lives. Equally, it does not seem correct to say that the later disabled child or adult is the person who is harmed by being brought into existence, since the only alternative for them would be not to have existed at all. Even if their life is so terribly poor as to be not worth living (a very rare case), it will not be true that they were harmed by not being aborted. To be made worse off by anything, one must already exist in some prior state that can be compared with the latter one. Of course, if the fetal interests argument is deficient for this reason, then it cannot be used to defend either FAA or SSA.
Because of the non-identity problem, the only way to claim that it is worse to bring new disabled persons into the world than to bring new women into the world, even when those women will suffer considerably under male oppression, is to argue that creating new disabled people contributes to an impersonally worse state of affairs. In other words, one must say that although it does not harm the individual brought into existence, it is of some negative moral value when new disabled people are brought into the world, and better when they are not. Suffice to say, I cannot imagine how one could convincingly make such an argument, and by reference to what values and goods the existence of people with disabilities is a bad thing (and of course, much can be said to the contrary).
-  ibid. 26 ibid 126.
-  27 Williams’s account is, however, notable for giving those arguments a fairly persuasive (ibid
-  144-50).
-  See, Christopher Gyngell and Thomas Douglas, ‘Selecting Against Disability: The Liberal EugenicChallenge and the Argument from Cognitive Diversity’  Journal of Applied Philosophy 2.
-  ibid 6.
-  With regard to the expressivist objection, for example, Williams considers the possibility thateven if women who choose to terminate their defective fetuses do not act on the belief that disabledpeople’s lives are inferior, ‘the combined effect of what they do will still be to promulgate such [objectionable] ideas throughout society’. His suggested answer to this is to say that the costs to women ofprohibiting selective abortion of defective fetuses are simply too high to be justified by the need to avoidreinforcing general prejudices against disabled people. In short, it is unfair to ask women to suffer for such a cause (Williams (n 24) 147).
-  This caveat about comparable burdens is in recognition of the fact that all things may not beotherwise equal. For example, one particular woman may face a far greater threat to her future wellbeing by the prospect of giving birth to a female child than another woman does by the prospect ofraising a disabled child. The prohibition of selective abortion as a means of combating prejudice mightonly be legitimate when the burdens it imposes on individual women are not unreasonably severe.Consequently, the suggestion here is not that it would be legitimate to prohibit all FAA in the interestsof social equality if the same considerations can ever justify SAA, but only that the cases ought to betreated similarly where the interests served by prohibition are equally strong, and the burdens on theindividual pregnant women are of comparable weight.
-  32 The ‘non-identity problem’ was originally formulated by the philosopher Derek Parfit as a challenge to a certain way of moral thinking which ties the wrongfulness of actions to their propensity tocause harm to identifiable individuals. See Derek Parfit, Reasons and Persons (Oxford University Press1984) part 4 generally.