Quoted from Textbook of Human Reproductive Genetics, chapter 13:

PGD within the“autonomy model”

According to what may be called the “autonomy model,” prospective parents are free to use PGD in order to select embryos on the basis of any characteristic they prefer, whether health related or not. Opponents argue that selecting for non-medical characteristics violates the autonomy of the future child as the child is reduced to an object of parental ambitions and ideals. But would embryo selection on the basis of characteristics that do not limit the possible life plans of the future child or that are useful in carrying out almost any life plan (“general purpose means”) really violate the future child’s autonomy? Should one not say that prospective parents undermine the ethical standard only when they deliberately try to direct the child toward a predetermined life? Anyway, the technical possibilities to use embryo selection for “superbabies,” whatever that may be, are regularly widely exaggerated in the mass media.

A paradigm case for the autonomy model in the current context is PGD/sex-selection for non-medical reasons. Sex selection for non-medical reasons is prohibited in many countries. From an ethical point of view, however, this is not evident [17]. Even though individual requests may stem from discriminatory attitudes or stereotyping views regarding the difference between boys and girls, it does not follow that sex selection for non-medical reasons is inherently sexist [18]. The fear that allowing it will result in a distortion of the sex ratio does not seem convincing either, at least not in Western countries, where a preference for boys is weak or absent. Moreover, the suggestion that sex selection for non-medical reasons will reinforce gender stereotypes to the detriment of children’s development and women’s position in society, are speculative at best. Since the conclusion must be that arguments against allowing sex-selection for non-medical reasons are weak, banning the practice may amount to an unjustified infringement of reproductive freedom. However, even if sex selection (limited perhaps to “family balancing”) may be acceptable in itself, a further question still concerns the proportionality of the means. Clearly, the use of preconception sperm selection technologies for this purpose (if safe and effective) is more easily justified than PGD [17].

A second case is PGD for “dysgenic” reasons. The paradigm case regards a deaf couple’s request of PGD in order to selectively transfer embryos affected with (non-syndromic) deafness. The couple may point to psychosocial and developmental risks of hearing children growing up with (two) deaf parents. Concerns include that (young) hearing children will have difficulties in understanding the implications of their parents’ disability and related behavior, that deaf parents will have only limited access to the experiences of hearing children, and that there is a risk of role inversion. Furthermore, applicants may argue that “deafness is not a handicap, but just a variant on the spectre of normalcy.” After all, deaf people have their own rich culture and their own (non-verbal) language. One can reasonably doubt, however, whether the “just a variant” view is tenable; after all, outside the microcosmos of the deaf subculture, deafness is a disability which causes a variety of serious and lifetime challenges. Tough deaf people still can (and usually do) live a reasonably happy life, selection for deafness is at odds with the professional responsibility of the reproductive doctor [11]. The couple’s relational concerns should be tackled by educational support and advice, not by “dysgenic” PGD. Interestingly, ongoing technology development may contribute to solving the current moral puzzle. Until now, cochlear implants are controversial, amongst other things, because their success is patchy. However, when the perfect version of the cochlear implant would become available in the future, parents will clearly harm a child they leave deaf. To select for a deaf child, then, becomes self-defeating.

People often use double standards for genetic vs. non-genetic interventions. For instance, in the above, there is talk of a violation of “the autonomy of the future child as the child is reduced to an object of parental ambitions and ideals”. Curiously, no one complains about parenting violating the same, despite the obvious attempt of parents to mold their children (not with too much success). Funny how parents are allowed to try to mold their children but only if it involves non-genetic means!

See also Nick Bostrom’s The Reversal Test: Eliminating Status Quo Bias in Applied Ethics.