In mid-2020 a paper was published in the medical journal, The Lancet, that caught my attention. The paper projected fertility, mortality and migration trends into the year 2100. The paper has some interesting results, including that the population size of 23 countries are projected to halve by 2100, and that China is projected to become the largest economy within 15 years, but the USA is projected to return to that spot before the end of the century. Unsurprisingly the paper also suggests that the global age structure will shift dramatically, such that the age pyramid is practically inverted.
However, I was particularly struck by the final two sentences of the paper:
“Policies that countries pursue today can alter the trajectory for fertility, mortality, and migration. Population size and composition are not exogenous factors for countries to account for in their planning, but rather outcomes that they can help direct.” (Vollset et al. 2020, p1304)
In the last few years, I've becoming increasingly interested in the area of philosophy known as population ethics. This field has largely sprung from the work of Derek Parfit, and particularly his 1984 book, Reasons and Persons. It deals with the question of how we ought to make decisions around actions that affect the number, identity and quality of life of people who will come to exist.
The questions that population ethics deals with are relevant to a lot of issues in public health and biotechnology. Some of these have received a lot of attention within bioethics—for example, the permissibility of abortion, or the question of whether it’s important to become a genetic parent—but others haven’t received as much attention as I think they should.
I was reminded of this recently when I read a new bioethics book, A Theory of Bioethics, by David DeGrazia and Joe Millum (which is available open access here). I really appreciate this book because it brings a level of philosophical rigour that is too often missing in bioethics. It also addresses some questions that I think have been neglected in bioethics, including non-human animal welfare, and population ethics. But although the book addresses one important population ethics question—the implications of determining the identity of a person who comes into existence—I was surprised by the omission of one of the core questions of population ethics: how (if at all) we ought to take into account, in our moral decision-making, changes in the size of a population. Or, to put it another way: how should we evaluate the ways in which our actions can affect how many people come into existence?
It surprises me that this question wasn't considered more in A Theory of Bioethics, given that one chapter, Chapter 10, is titled ‘Creating Human Beings’. I think this is a shame, because it's relevant to a lot of the questions the authors consider, including whether we ought to subsidise assisted reproductive technologies, the permissibility of abortion, and how to evaluate policies that affect fertility on a large scale. I also think the ‘Dual-Value Theory’ that DeGrazia and Millum present in their book might provide a good framework for dealing with some of the difficulties of determining how to value different sized populations.
In this post I’ll share some thoughts I had on reading this chapter, with the goal of showing why it's important to take the core questions of population ethics seriously in bioethics and health policy. I will make a lot of references to the chapter, and the book more broadly, but I don't think you need to read it to follow this post.
Valuing reproductive technologies
In Chapter 10 of their book DeGrazia and Millum first argue for a negative right to procreative autonomy – that is the right to non-interference with one’s reproductive choices. They argue that this grounds a right to access contraception, and a right to choose whether to continue or terminate a pregnancy, at least up to a certain gestational age.
The authors then address the question of whether a person has a positive right to procreative autonomy, that is, a right to access assistance to procreate. They argue that people’s interest in procreation is better thought of as an interest in parenting, that is having the opportunity to raise a child. They then suggest that this interest in parenting would be better met by enabling people to adopt rather than create a new child, since adopting a child who needs to be cared for is more important than creating a child who doesn’t yet exist and doesn’t yet have any needs to meet. However, the authors acknowledge that in most countries adoption is actually very difficult and there are few infants in need of adoptive families. There are more older children in need of families, but adopting such children comes with greater costs to parents: they miss the opportunity to form bonds with the child in their important early years, and they are likely to face a much more difficult parenting situation with these children. For this reason, the authors suggest that adopting rather than procreating is not morally obligatory.
However, they don’t really return to the question of whether, given that the interest in parenting can’t realistically be met by adoption, we ought to provide access to assisted reproductive technology. They do note that the reproductive rights of many more people can better be met by providing access to contraception and family planning services. This is true, and I agree that ensuring this access ought to be a priority for health care provision. However, it’s not clear why the only goal ought to be “meeting reproductive needs”, rather than a more general goal of providing cost-effective health interventions. In the authors’ Dual-Value Theory, both well-being and rights ought to be considered. So, it seems consistent with this approach to also consider the well-being effects of procreation-related policies, in addition to their effects on rights.
This raises a question of whose well-being ought to be considered, and specifically whether we should value the well-being of a child created using reproductive technology. An act of procreation usually has profound effects on the well-being of parents, but it also has important well-being effects for the person created. The authors highlight this when they write:
“… whether the act of procreation has positive value presumably then depends on the fate of the person so created.” (DeGrazia and Millum 2020, p283)
However, the potential positive value of an act of procreation for the person created receives little attention in the rest of the chapter’s analysis. The authors do consider the possible negative effects of procreation for a person created, when they raise (but don’t defend) David Benatar’s view that it is wrong to bring people into existence given that they will inevitably experience suffering in their life. However, the arguments against Benatar’s view, the potential positive value of a person coming into existence are not mentioned.
The procreation asymmetry
It’s possible that the authors are assuming what has been labelled by Jeff McMahan as the procreation asymmetry – in McMahan’s words this is the view that
“… the fact that a person's life would be worse than no life at all (or "worth not living") constitutes a strong moral reason for not bringing him into existence, the fact that a person's life would be worth living provides no (or only a relatively weak) moral reason for bringing him into existence.” (McMahan 1981, p100)
On this view, one ought not to consider the positive benefits to an individual when evaluating their creation. However, justifying this procreation asymmetry is a complicated task. If we want to consider the interests of people who come into existence with a miserable life, then consistency suggests we should consider the interests of people who come into existence with a happy life. We might think that avoiding harm is more important than benefiting a person, but that doesn’t usually mean that we have no reason at all to benefit a person, or that there is nothing morally good about benefiting a person. A harm-centered approach might justify it being more important to avoid creating a miserable person than it is to create a happy person, but not necessarily that there is no reason to produce a happy person.
Some worry that if we accept that we have some reason to create a happy person, then this will entail an obligation to procreate, and perhaps even an obligation to produce as many people as possible. However, DeGrazia and Millum's Dual-Value Theory provides a way to resist this conclusion. As well as valuing well-being, it also values rights. With respect to well-being: we can accept that consideration of well-being gives us a reason to bring a person into existence, but we also need to acknowledge the negative well-being effect that procreation can have on parents and other existing people—for example, if prospective parents don’t think they are in a situation to raise a new child. And, of course, perhaps more importantly we also need to consider the right to reproductive autonomy which the authors defend in this chapter. This allows us to accept that there is some value in producing a person with a happy life, but that this is not obligatory.
When it comes to the permissibility of abortion, taking this approach will lead to broadly the same conclusions as the authors come to, that is, that abortion is usually permissible. The authors argue for the permissibility of abortion, by noting that prior to the onset of sentience a fetus doesn’t fit their definition of a being with moral status. In addition, they argue that on their definition of the harm of death, a pre-sentient fetus would not be harmed by death because it has no psychological connections to its future life. This ‘psychological connection’ criterion also applies to sentient fetuses, and suggests that abortion of a sentient fetus may be permissible if there is an adequate reason to do so. They suggest that one such adequate reason is a person’s interest in avoiding the significant burden of becoming a parent.
However, if we accept that there is some value in creating a happy person, then this implies that there is something wrong in aborting a pre-sentient fetus, even if this is outweighed by the well-being and rights-based reasons to allow the abortion. This 'something wrong' raises the issue of so-called “impersonal effects”, because if there is a wrong that is committed by aborting the pre-sentient fetus, it is a victimless wrong, as the pre-sentient fetus isn’t a being with moral status.
Identity and impersonal effects
This intersects with the chapter’s discussion of the non-identity problem. First described by Parfit in Reasons and Persons, this is the problem of how to consider decisions that cause a person to exist, but at a lower level of welfare than another person may have enjoyed had a different decision been made. An example may be a parent’s choice to conceive a pregnancy where there is a risk of infection that would cause congenital abnormality, such as infection with the Zika virus, an example which the authors review closely in the chapter. Zika virus causes congenital abnormalities including microcephaly and a greater risk of child mortality. In many countries couples are advised to delay attempting to conceive until months after returning from a Zika-endemic area. If they don’t follow this advice and conceive a child who is affected by congenital Zika syndrome then this specific child could not have existed without the syndrome. So, their action has not wronged this child. The authors argue that in this situation, it is not wrong to fail to create the best life possible, as (setting aside the effects on people other than the child) no one is harmed by this action, it is a victimless wrong.
However, the authors' intuitions are different when we consider other instances of the non-identity problem, such as Parfit’s case of Depletion (also from Reasons and Persons). This describes a scenario that we as a population face currently: whether to conserve or deplete the earth’s resources, with the assumption that choosing the option of depletion will result in lower well-being for future generations. Whichever option we choose will also have the effect of changing the identity of the specific individuals that make up the future generation, as presumably the changes in behaviour that result from depletion or conservation will alter who procreates with whom and at what time. So, just as the child with congenital Zika syndrome can’t blame their parents for whatever harms the syndrome brings, so too future generations couldn’t blame us for choosing depletion, as had we chosen conservation those people wouldn’t have existed.
Unlike in the individual case, most people think that we would be doing something wrong if we were to choose depletion, even though this would seemingly not result in any identifiable victims. DeGrazia and Millum provide three possible responses to this situation: 1) to bite the bullet and accept that we wouldn’t be doing anything wrong to choose depletion, 2) to accept that it can be wrong to create a world with less well-being even if this doesn’t result in any identifiable victims, or to put it another way, there can be morally relevant effects that don’t harm any person, which we can call impersonal effects, and finally, option 3) is to differentiate between two types of non-identity cases. The authors favour this final option, and draw on work by Melinda Roberts and David Wasserman to support their view. For reasons of space I won’t review their rationale here. But it is worth noting that the authors also accept response 2) at least to some extent. On their view impersonal effects matter, although they say that they don’t think they carry significant moral weight.
However, in their evaluations of two cases the authors seem to suggest that these impersonal effects can sometimes be decisive. They provide two examples to illustrate when there is a sufficient reason to have a child with a worse life than is possible. Firstly, they suggest that a couple who choose not to delay a pregnancy that would result in a child with a disability do not act wrongly when doing so might mean they miss their opportunity to become parents. But in this case, there isn’t an option of creating a child with a better life. The option that couple chose was the one that led to the greatest well-being. If they had chosen otherwise there would be no child. On the other hand, DeGrazia and Millum suggest that a couple who don’t avoid conception of a disabled child because they simply can’t be bothered getting contraception would be at fault, as this isn’t a good enough reason to not produce the child with the better life. The rationale they give for this is that it is worse to create a world with more suffering. So, it seems that a concern for impersonal effects (how much suffering there is in the world, rather than the suffering of a particular individual) is driving the judgment here.
Summing up, I think we should consider the well-being of potential people when we think about the ethics of creating human beings, but that this is one consideration, that needs to be taken alongside consideration of the well-being and rights of other people.
How we ought to consider these effects together is unclear. Should we consider the potential well-being of possible people as equivalent to well-being of existing people? How should we account for the effects of reproductive autonomy on well-being? And how should consider rights in all of this? Should rights always trump well-being effects?
I think it’s plausible that we should consider the well-being of potential people the same way we consider the well-being of existing people, but that certain rights ought to take priority over well-being considerations. This approach could allow us to value potential people but also not imply that we have an obligation to procreate or produce as many people as possible.
There is a lot to think through in this. From a philosophical perspective, there is much more that can be said about population ethics and in particular the procreation asymmetry. And indeed, there are a lot of smart philosophers who have thought and written about this issue. In this post I’ve barely scratched the surface of this literature.
There are also tricky empirical questions. What should we expect will be the effect of adding more people on the average level of well-being experienced in a population? In the book, DeGrazia and Millum note that creating more people could have detrimental effects on climate change, as there will be more people consuming resources including energy. But elsewhere economists have warned of the potential harms of population stagnation. It’s also unclear how we should account for the positive well-being effects of reproductive freedom. This includes both the direct effects on individuals from having more control over their lives, and the societal effects that come from having more people (and especially more women) freed up to pursue other projects.
For all these reasons and more I definitely don’t want to suggest that we should be pursuing a policy to increase the population as much as possible, as quickly as possible. But I do think it’s important not to completely discount the value of creating new people.
Why is this important?
I think DeGrazia and Millum have done an important service to bioethics by highlighting the relevance of population ethics to the field. But as well as the issue of non-identity, the issue of how to value potential people is one that is becoming increasingly necessary to deal with.
This bears on choices around how to value assisted reproductive technologies, like IVF, and on other policies that alter fertility. One example is climate change policy. Not too long ago, I was talking to a climate scientist who in their most recent work had produced different sets of results taking different assumptions about the value of population size, not knowing which results should be taken as correct. I’ve had a similar conversation with an economist working on climate change policy. Economists are now including changes to mortality rates in their estimates of the social cost of carbon, but don’t know whether or how to include changes to fertility rates. More theoretically, how we value potential people will also have implications for how we value existential risk reduction.
Thinking back to The Lancet paper I mentioned at the start of this article, one thing is clear: we have the power to influence how many people will come to exist. What we should do with this power is unclear, but I don’t think the answer is to ignore it.
If you have any feedback on this post or notice any errors, please let me know through this anonymous form. I'm grateful to Robert Long for helpful edits on an earlier draft of this post.
 This is another topic that I would like to write a post about, but a lot of people in bioethics don’t have training in ethics. I think this lack of philosophical rigour is a problem for the field, as do some others.
 I’m skipping over it because this post is already quite long, but this is essentially the Time-Relative Interest Account of death, developed by Jeff McMahan in his 2001 book, The Ethics of Killing. (A handy hint - if you read this on the train with title clearly visible, you might end up with more free seats around you.)
 This issue is also explored in a paper by Doolabh et al., which as well as giving a nice introduction to the non-identity problem provides the results of a survey of opinion on this problem with members of the public.
 This website offers a starting point for readings on population ethics. Some other papers that I’ve found interesting are: Asymmetries in the morality of causing people to exist by McMahan, Future People, the Non-Identity Problem and Person-affecting Principles by Parfit, and Conditional Reasons and the Procreation Asymmetry by Johann Frick.