Selective Abortion

Preface
This post is largely culled from an old email I wrote. (Now it can be ignored by the public.) Please regard any deficiencies in reasoning as youthful indiscretions - but don't hesitate to crucify me for them.

Body
I once faced the following debate case: There's an average American family with one young kid. The wife is pregnant, and the couple is certain that the baby will be seriously autistic. Should they abort?

I unhesitatingly picked yes. I maintain that position, but it got me thinking about how the decision to carry a fetus to term should be regulated. To begin with, it's clear to me that abortion should be mandatory in some cases. For instance, one shouldn't be allowed knowingly to have a kid with infantile Tay-Sachs. People often seem to lose sight of the fact that child-bearing is far from a self-regarding action. I also think it's unquestionable that abortion should be discretionary otherwise. To me, the interesting, and intertwined, issues are (1) what information parents should be allowed to have access to in making the decision to abort (information is the engine of discretion); and (2) how robust the mandatory provision should be (should it encompass AIDS? Deafness? Dwarfism?).

My initial inclination was that parents should be able to learn whatever they want about the fetus. I suppose I had three reasons for favoring fully-informed discretion. First, because the parents will presumably raise the child, it's better for both them and the child if the child doesn't possess traits that the parents find objectionable. Second, it's best for parents to be able to deal with cases at the cusp of the mandatory provision. Because it's so severe to require an abortion, the mandatory provision should err on the side of leniency; doctors should be required to satisfy stringent criteria, subject, of course, to state review. Consequently, the provision won't apply to fetuses that almost all pro-choice parents think should be aborted (e.g., deaf ones); these parents should be able to step in. Finally, given that it's impossible to bar access to all fetal information, there are reasons for favoring equal access. This isn't an argument about a black market in fetal trait screening (not that it wouldn't exist). What I mean is that parents inevitably have some idea of how likely a fetus is to possess a trait, depending on the nature of the trait (e.g., two genetically deaf parents have a good idea of the likelihood that their offspring will be deaf, whereas fat parents have a slimmer notion of how fat a given child will be inclined to be). Why allow the vicissitudes of genetics to determine the extent to which parents can select for or against certain traits?

Now I'll attempt to argue against my intuitive stance. In response to my first reason above, consider that the harms of parental expectations would be exacerbated if parents could engage in selection, however crude, of children's traits. For an extreme example, parents who abort whenever the child would be insufficiently athletic (assuming this could be screened for) might put even more pressure on their kids to be good at sports. Another plausible concern, for what it's worth, is that kids with particularly bad traits might resent their parents for not aborting them. Turning to the second point - the heart of the debate - maybe it's simply bad if parents can be selective. Maybe the world is better off with "healthy" numbers of short, socially awkward, and depressive people (especially if these traits are correlated with beneficial ones, such as hyperintelligence and being Woody Allen). One may retort that children shouldn't be born worse off for the good of society, but the implicit parallel to punishing innocent people doesn't necessarily hold. Children who would have been aborted had their parents been aware of their impairments aren't being made worse off, and their lives have some value. The point is that personal utility is largely relative. Sure, some conditions, such as chronic pain, are just bad (and are largely addressed by the aforementioned mandatory provision). But other conditions are only bad if one acquires them, not if she's born with them (e.g., being of slightly below average intelligence, or even, arguably, deafness). In light of this distinction, the issue can be framed as whether it's acceptable for parents to be able to shift and/or compress traits' bell curves via selective abortion. Shifts are often pointless because of relativity (e.g., if everyone were an inch taller, the only benefits would be minor and incidental, and may be offset). Of course, this point is harder to defend when it comes to traits for which meaningful absolute gains are possible (e.g., suppose everyone were smarter). Regarding compression, imagine a world in which almost all men are between 5'11" and 6'3", or in which almost everyone's IQ is between 110 and 130 (apparently a fair amount of people believe in the pop psychology notion that too much intelligence necessarily precludes healthy social development). Compressing the range of a trait can be sterilizing and risky. We can be bereft of meaningful diversity - and its often unforeseeable benefits. Perhaps this would be the primary consequence of widespread, fully-informed selective abortion.