Surrogacy is occasionally a target of criticism from the online right. The idea that a mother would have the baby she just gave birth to taken from her and given to a couple—especially a gay couple—is particularly repugnant to some. It is often argued that this is an injustice for the biological mother, even if she voluntarily agreed to surrogacy. Another common argument is that it is immoral to buy and sell babies or the service of pregnancy. While a complete defense of surrogacy would address these arguments, I want to narrow my focus on arguments about the child's welfare. These arguments typically emphasize the need of a child to have a mother.
Those with a pro-surrogacy inclination may want to dive into an empirical argument about how a gay couple can be good parents or about how the outcomes for children born via surrogacy are not inferior. While the empirical arguments are worthwhile, there is a much more concise argument that severely undermines the case for opposition to surrogacy. If you are concerned about the well-being of a child who never meets his or her biological mother, then you need to seriously consider the alternative scenarios.
If surrogacy for gay couples were made illegal, the counterfactual is not that the surrogate mother keeps the baby. The correct counterfactual is that the child would have never been created in the first place. Whatever negative aspects you believe there are to never meeting one’s biological mother, it is quite difficult to imagine that never having been born is better. While arguments about the importance of motherly love are reasonable, it is hard to defend the idea that life is not worth living without a mother. An extreme adherent to this view might see no need to save the life of a child whose mother has died during labor.
Intuitions on surrogacy are severely distorted if one considers their own mother or own child. Children born via surrogacy to gay parents would likely have different intuitions. Perhaps they would find it extremely distressing not to have their second father. Surely, most would consider their life worth living.
The most reasonable criticism of surrogacy from this angle would be if women prevented from being surrogates had additional children, and the quality of those additional children’s lives was better. The argument would be further bolstered if the would-be surrogate’s happiness with her child was greater, too. Such an analysis is much more difficult. My personal intuition is that this is probably highly unlikely and surrogacy wins from a cost-benefit perspective.
The key point regarding surrogacy is that considering the counterfactual is important. Another apt example of this would be the case of in vitro fertilization (IVF)—especially the earliest instances of successful use of IVF, which were met with skepticism. It is easy to understand why someone would have been concerned about IVF being unsafe before the first IVF baby—Louise Brown, who was born in July of 1978.
However, it is important to consider the counterfactual. If the use of IVF had been substantially delayed, Louise’s parents would have likely never even had a child as they had been trying to conceive for 9 years already. Surely, a potentially less than perfectly healthy child is better than no child at all. The exception would be if there were a strong expectation that the negative side-effects of IVF were so extreme that we could expect Louise Brown’s life not to be worth living at all.
Concerns about safety are rather reasonable, but they need to consider the alternative. It is reasonable to want medical interventions to have undergone thorough testing, but if it means that someone does not even get to exist, then the precautionary principle is being taken too far. It is ethical to permit people with a terminal illness to use experimental drugs in the hope of saving their lives if they have given fully informed consent.
A high-profile case where this analysis does not apply is that of Chinese researcher He Jiankui—who used CRISPR-Cas9 with CCR5 as a target in an attempt to confer HIV resistance on the twin babies Lulu and Nana. The use of CRISPR was unnecessary to bring about the existence of Lulu and Nana, as the couple could have transferred the embryos without editing. Had the couple been completely committed to only having Lulu and Nana if they could gene-edit, the analysis would be different. One could still have a coherent view that this preference from the parents was highly questionable, and He was unethical in facilitating the creation and transfer of gene-edited embryos.
A preference for a genetic trait so strong that one would not otherwise conceive a child is not unreasonable. It would be defensible if a couple with a family history of schizophrenia were willing to conceive a child only if they could ensure that the risk was beneath some threshold. While this was recently impossible, technical and scientific advances have enabled the use of polygenic risk scores in preimplantation genetic testing. One of the most powerful arguments in favor of polygenic embryo screening is the fact that cases could exist in which couples would not otherwise have a child without polygenic screening. The cost-benefit analysis shifts strongly in favor of having the child after screening. This is another reason why blanket condemnations on polygenic embryo screening are misguided.
Perhaps most controversially, I will now defend the creation of the first IVG baby. IVG—in vitro gametogenesis—is the creation of gametes (i.e., sperm and eggs) in a laboratory setting. While we are currently able to convert somatic cells—such as blood or skin—to what are called induced pluripotent stem cells (iPSCs), we are not yet able to convert human iPSCs to derived gametes. The creation of derived gametes has been achieved in mice and may soon be possible in humans. Like surrogacy, this is an important discovery for gays and lesbians as it would help facilitate the creation of children from two same-sex parents. Hopefully, this contributes to the defense of IVG by progressives, especially seeing as such an idea will certainly be repulsive to many conservatives.
Once IVG has been achieved in the laboratory, the next hurdle will be to create a human being from derived gametes. Mice have already been created from derived gametes; some mice were even born from two male mice. Even if IVG were to be demonstrated to be safe in mice, many people would be nervous about creating the first IVG baby.1 There could be unforeseen epigenetic or transgenerational effects. Especially concerning is the fact that somatic cells would accumulate more deleterious mutations compared to what eggs would normally have.
Once again, we should consider the counterfactual before we condemn the first use of IVG. If a couple were incapable of creating a child in any other way, then the relevant moral comparison would not be between an IVG baby and a naturally conceived baby. The relevant comparison would be an IVG baby with no baby. If IVG appears safe enough such that the life created would be worth living and not a net negative to the world, then it seems perfectly defensible to create the first IVG baby. Surely, after this baby has been created, few would say that the life is not worth living—a judgment best given by the first IVG baby once they have grown older.
While unforeseen negative side-effects are surely concerning, there are some fortunate benefits to IVG. Namely, it would enable the creation of large batches of embryos. This is beneficial because massive batches of embryos paired with inexpensive polygenic embryo selection would enable very large returns in physical and mental health outcomes over the expected baseline. Even with a reduced baseline due to higher mutational load, expected returns would likely make this an attractive proposition. This is especially true if there are a handful of babies born without any apparent side effects.
With batches in excess of 1000, the tradeoff would then involve 20-30 IQ points or perhaps more than a decade more years of life. Even if there were an apparent downside to being an IVG baby, the use of polygenic selection would likely more than account for that. As the first IVG children grow older, the concern of some yet-to-be discovered side-effects will likely dwindle. Seeing as the benefits are so immense, intense examinations of these children would be of the utmost importance. Despite skepticism, surely many parents will believe the scales tip in favor of IVG and privately use it if it does in fact prove safe. Thankfully, the desire of parents to better the lives of children is strong, and so I am optimistic that IVG will not be left unused for decades.
Bioethics and population ethics are both very difficult topics as many people have strong moral intuitions and disgust responses. In many cases, it is more straightfoward to sidestep in-depth empirical discussions and make a direct comparison with the most relevant counterfactual. I believe this approach makes it apparent that surrogacy, IVF, and creating the first IVG baby—under certain plausible assumptions—are all ethical. To argue otherwise in many cases amounts to arguing that these practices have or will produce people with lives not worth living, something that is surely not generally true.
See this discussion between Steve Hsu and Aella on the ethics of IVG (bolding mine):
Steve Hsu: I see. I see. Yes. So, using pluripotent skin cells to make eggs, kind of de novo. The situation with that is the following. So, labs have gotten it working quite well in other animal species like rats. And I believe actually maybe there's some primate species where they've maybe got it working. Nobody has reported getting it working in humans yet. It's a very complicated subtle wet lab thing where if you like just change the procedure slightly you could screw it up.
Like for the rat stuff, for a long time It was just a few Japanese labs that really had it working well, and maybe that spread a little bit. Now there's some startups that are actually trying to do it. But one thing to consider is the following: do you really want to be the first to test this? Because how, although it appears that the rats produced and monkeys produced by this are normal, who knows what effects there will be on your child from having been, having been produced from an egg made in this way. We don't really know.
And so my prediction for that industry is, since we already have tried and true IVF technologies and stuff, which don't require that, it'll take a while for that to get normalized and adopted at scale. That's my thought on it.
Aella: So I think you were saying in order to, what, what would be the method to decide like an egg produced originally from a skin cell is just as good as one produced the other way, the original way. In a way like this question is still not known for IVF. You know, the oldest IVF person is under, I think under 45 or around 45 now.
Steve Hsu: In fact, the oldest American produced the first American baby produced in IVF turned 40 just recently. She actually works for Genomic Prediction. She's a customer. She's a patient advocate for us. And the British baby who's a little bit older Is in the UK.
But still like you like in principle one could say, oh, you know when these IVF babies turn 65, there's something really terrible is going to happen to them. Now, you might say, oh, mechanistically, that's implausible because like, what's the difference? you know, and whatever effects were happening were happening, like when the thing was only like 100 cells big. So really, by the time they're 60 or 70, it's all got to kind of even out. But in principle, people are still monitoring like the oldest IVF babies to see if they can detect any, you know, differences between those babies and babies produced the old fashioned way.
Okay. If you introduce something really radical like, oh I made the egg that became you out of a skin cell, and wow we found out that cancer rates are like 10x higher, 3x higher for that population, but we only found that out when people got to be 60 or 70 years old. That that's that would be the concern that people have yeah.
And you can say like, oh, but we've been studying these rats and monkeys that were made through this technology, you know, and the rat in the rat case, many generations have been studied in the monkey case, not so many actually. Probably not even one. But that would be the concern.
And so, like, the question would be, like, what's your reason? Like, if you're a woman who missed her fertility window and wasn't able to have any kids and you're 55. And but you still want to do it and you're going to use the skin cell for sure. I think they're going to be people because the desire to reproduce in some people is so strong that you will have some cases like that. But I think as a fraction of the utilization relative to the utilization of standard IVF I think these new technologies are going to be tiny for a long time to come and in for IVF in particular like in some countries, like Scandinavian countries, or I think maybe Israel, I think almost 10% of babies now born are actually born via IVF.
So it's a non-trivial fraction of the entire birth cohort each year.
Aella: Yeah. I think like for me, it's like a trade off, right? Like the benefit of IVG is that you get to make a [s***] ton of embryos and then test a [s***] ton. And if you can, if we had the technology to select for IQ or happiness or whatever, then like the potential to be truly brilliant, to really introduce like truly brilliant people into the population, which is, what would be a magnificent benefit. Like, to me, it's probably worth the risk. I would rather want to see because it's like we're not just risking, like maybe We make babies that get cancer. We're also risking it not having a universe where we have a bunch of brilliant people I don't know.
So I hope it gets made pretty soon.
Steve Hsu: Yeah, you know, your, your take is the rationalist take which, you know, I'm obviously sympathetic to. But then that gets me called a eugenicist and all kinds of nasty names and stuff like that. […]
> Even with a reduced baseline due to higher mutational load, expected returns would likely make this an attractive proposition.
Furthermore, mutational load is probably a nothingburger: https://www.sebjenseb.net/p/taking-mutational-load-seriously
Well said.