With that out of the way, I wish to suggest a division of our notion of "eugenics" into three categories: pro-natal eugenics, which aims to increase the number of people being born with preferable genes; anti-natal eugenics, which aims to reduce the number of people being born with less-preferable genes; and improvement eugenics, which aims to improve the genetic quality of people who are going to be born anyway. An example of pro-natal eugenics would be providing financial subsidies for high-IQ couples to have children; an example of anti-natal eugenics would be compulsory sterilisation of people judged to be defective in certain ways; an example of improvement eugenics would be screening embryos for disease among people undergoing IVF treatment.
These different kinds of eugenics ought to be assessed differently. My key thesis here is that improvement eugenics is clearly desirable, pro-natal eugenics is likely to be anti-egalitarian but that the good consequences may well outweigh this, and that confusing these with anti-natal eugenics is responsible for most of our worries about eugenics. (I'm not going to take a strong position on whether anti-natal eugenics might be overall justified, but it seems far more problematic than either of the other kinds).
There is a risk with any of these programs that policy-makers will seek to promote not good traits but traits which they personally like - for example, particular colours of skin. We should acknowledge this danger, should fight against all such misapplications of eugenics, and may find that the risk of abuse is to high to practice any eugenics. But I maintain that these practical issues are irrelevant to the in-principle-acceptability of certain kinds of eugenics.
When we talk about ways to improve outcomes for people who will exist anyway in ways which don't involve genetics, no-one bats an eyelid. Controls on lead emissions are obviously desirable. Education, insofar as it represents real improvements in people's capabilities rather than just a form of signalling, is similarly desirable. The only question, then, is whether the fact of these changes being genetic rather than through other mechanisms makes a moral difference.
It does introduce some extra reasons to be concerned, to be sure. Genetic changes are rather harder to reverse than many other kinds of change: if it had turned out that we were wrong about lead and that it was in fact vital to children's development, we could start pumping it into the air and would within a few years fix much of the damage caused; if it turned out that an incident of gene editing had significant negative consequences, this would take longer to correct and would require significantly greater resources, if it was even possible. But this does not affect the case, on the level of pure principle, for improvement eugenics.
Some people object that by meddling with genes, we are "playing God". Given that I don't believe in any kind of God, at least in the conventional sense, I'm not inclined to take this kind of argument seriously. Besides which, such arguments seem woefully underspecified. What is that makes fixing people's genes blasphemous, but throwing a ball not blasphemous? Both involve meddling with the world in certain ways which might happen to be either in accordance with or contrary to God's will. I'd be happy to have this discussion with a serious religious thinker willing to supply such a condition, but in the absence of such an interlocutor I feel the attempt would be a waste of time.
Perhaps changing someone's genes involves some kind of interference with their autonomy. OK, but it can also improve their autonomy if it leads to higher intelligence, conscientiousness, or similar. Moreover, it's highly unclear why we should take their natural set of genes as the moral default from which any deviation must be justified.
Ultimately, it's hard to see why we shouldn't edit out things like hereditary diseases from the human genome, unless there are significant side-effects of doing so. Is it right to save life, or to kill?
This is more problematic than improvement eugenics. Most obviously, it's likely to involve anti-egalitarian transfers of resources and welfare, since the people we would be attempting to incentivise to have more children would in many cases be those who already have plenty. (Perhaps the solution would be, rather than rewarding high-IQ types for having more children, punishing them for having fewer children? But even if this is judged worthwhile when intelligence we wish to encourage, it becomes rather less palatable when trying to encourage greater procreation by people with genes that lead them to be more pro-social than average, or other things we view as virtuous).
That said, I think in general this ought not to be much more controversial than improvement eugenics. If you accept my arguments that people benefit from existing, and you think that certain people create net positive externalities for the rest of society (and would continue to do so on the margin if there were more of them), then why would you not want more of those people? Yes it has certain inegalitarian aspects, but any good Rawlsian should recognise than in the end we all benefit.
This is the bad boy. This is the kind of eugenics responsible for giving eugenics in general a bad name, the kind of eugenics used to justify forced sterilisation of despised minorities.
When considering any kind of anti-natal eugenics aimed at abolishing a condition X, there are two questions to be asked: (1) what does X mean for the quality of life of the person who possesses it? (2) Do people with X tend to make the rest of society worse off?
If the answer to (1) is that X usually makes people's lives not worth living, as with certain medical conditions, then we do not need any kind of eugenic principle to justify preventing people with condition X from coming into existence: we need only a sense of mercy.
If the answer to (2) is no, that they simply end up in a worse condition than the average member of society - so what? Let these people live, let their parents have full reproductive freedom!
The complicated cases come when a person is fully capable of having a life worth living, but would impose costs on society in doing so. Sometimes these costs will be concrete, such as those who are in important ways disabled at a young age and so require another person to act as a full-time carer. Sometimes they will be harder to detect, such as the stuff Garett Jones writes about. My tentative inclination is to think that some anti-natal eugenics may be permissible in such cases, but I cannot claim to have thought this through in any great detail. Moreover, any interests society may have in avoiding these costs must be weighed against various interests - in particular procreative interests and bodily autonomy - of the would-be parents of children with condition X. Paying criminals to be sterilised is probably acceptable, mandatory sterilisation is probably not.
Eugenics gets a bad rap due to the genuinely reprehensible things which it has been used to justify. However, eugenic interventions aimed at improving the genetic quality of people who will be born in any case and/or at increasing the fertility of people with desirable traits are in principle morally acceptable - though we might nevertheless have justified worries about the practicalities of such programs.