PEA Soup’s public philosophy section ‘Soup of the Day’ is pleased to introduce a new series with the online magazine Aeon, in which we will be hosting critical responses to Aeon articles written by professional philosophers for a more general audience, intended to kickstart a broader discussion. We hope that this will provide a fruitful forum for academics and the interested public to engage on contemporary issues together.
The first piece in the series is from Tim Campbell (Institute for Future Studies) on the ethics of population decline, responding to the Aeon article ‘The vanishing of youth’ available here: https://aeon.co/essays/population-decline-will-rob-us-of-vital-social-force-youth.
Hi Pea Soup readers! I’m glad to kick off the discussion.
I read Victor Kumar’s “The Vanishing of Youth” with great interest. Kumar argues that declining fertility will lead to a shrinking population in the next century, and that the resulting loss of youth will have serious costs: slower economic growth (even degrowth), reduced innovation, and the erosion of moral, political, and cultural life. He identifies causes of low fertility: delayed marriage, urbanization, weaker family networks, the burdens of parenting, and society’s failure to value parenting. Finally, to slow fertility decline, Kumar proposes “progressive pronatalism”: policies like paid leave, subsidized childcare, fairer division of parenting, more affordable housing, and experimentation with cooperative parenting, artificial wombs, and AI nannies. Some of these policies have benefits beyond fertility.
The article contains three types of claims: (1) empirical—claims about when population will peak and causes of low fertility; (2) evaluative—why a shrinking population would be worse than a steady (or less slowly shrinking) one; and (3) normative—why we should pursue pronatalist interventions.
I am unsure whether to endorse progressive pronatalism, since I’m stuck on the evaluative part. I remain uncertain how to weigh costs and benefits of a shrinking population. To do this, fundamental questions must be addressed: How should we value population? What makes one world better—or worse—than another?
Some of the goods Kumar highlights, such as growth and innovation, seem merely instrumental—they matter only insofar as they promote (or diminish) what is intrinsically valuable. But what is intrinsically valuable?
I’ll suggest some answers. I begin with background on global population trends, then address the evaluative questions. My framing is in terms of population ethics, which concerns valuing populations in which the number of people and their welfare may vary. I then present one reason why a shrinking population would be worse than a steady (or more slowly shrinking) one, supporting Kumar. Finally, I consider two possible benefits of a shrinking population—one concerns animal welfare, the other technological risk.
One important issue is what timeframe we should consider when evaluating population decline. Because Kumar seems to be concerned with the value of our world over the time when population declines, much of my discussion will focus on that. However, we can also ask what difference population decline, in contrast to population stabilization or slower decline, would make to the expected value of the entire future. When we ask this question, I suggest, somewhat speculatively, that animal welfare becomes much less important, and technological risk becomes the main focus.
My discussion does not conclude that a shrinking population will be overall better or worse than a steady one. My main aim here is to explore fundamental questions rather than resolve them.
Empirical Background: Why population decline might happen sooner
Kumar asserts that global population will peak late this century before declining. This is consistent with the UN World Population Prospects (2024), which projects ~10.3 billion in the mid-2080s, followed by slow decline.
But this forecast assumes that in countries with very low fertility (<1.5 children per woman), fertility will rebound to ~1.75. Some demographers question this. In an important Lancet article, Vollset et al. (2020) observe that countries like South Korea, Taiwan, Singapore, Italy, and Spain have sustained very low fertility with little recovery. They use a demographic model from the Institute for Health Metrics and Evaluation (IHME), which I think is somewhat more realistic than that of the UN. It predicts a peak of 9.7 billion in 2064, falling to 8.8 billion by 2100—decades earlier than the UN projection. This indicates the decline is more urgent than Kumar suggests; he may live to see the peak.
To focus discussion, I consider two possible scenarios: The Shrinking World and The Steady World. Each begins in 2064, when IHME projects the peak, and ends in 2200. Hence, these two “worlds” are not whole worlds, but just time-slices of worlds. (More on this later.) The 2200 end point is somewhat arbitrary; one could try to imagine the world out to 2300, or even later. But then it gets harder to tell how demographic trends might change.
The IHME forecast runs only until 2100. By then, it projects the “total fertility rate,” or TFR—to fall to about 1.66 children per woman, well below the replacement level of 2.1. (The TFR is simply the average number of children each woman would have over her lifetime if current birth rates stayed the same.) However, this estimate is derived after adjustments for uncertainty. In my description of The Shrinking World, I rely on IHME’s raw reference data, which puts TFR in 2100 at 1.79. (IHME has made their data and methods publicly available: IHME Global Population Forecasts 2017–2100.)
To imagine what happens after 2100, I extend this forecast. Here, I make some admittedly speculative assumptions, but I don’t think they are very unrealistic. I assume that TFR keeps slipping, but slowly, ending up at 1.55 by the year 2200. I also assume small improvements in life expectancy each generation. In doing this, I adopt the IHME’s own modelling assumptions and their life tables, which are tools demographers use to describe how many people at each age are expected to survive to the next age. When you put the pieces together, the result is a steadily shrinking and aging population.
The Steady World, on the other hand, is just a hypothetical benchmark against which we can evaluate population decline. In the Steady World, population size also peaks in 2064 at 9.7 billion, but then stabilizes. The birth rate magically jumps back to replacement level in 2064 and then adjusts so that every year, the number of people born exactly balances the number who die. In other words, the Steady World “freezes” humanity at its peak size, keeping population size constant until 2200.
The Shrinking and Steady Worlds are compared in the table below. Three specific years, 2064, 2100 (a reference point) and 2200 are represented for each world. For each of these years, global population size is given as well as population size (and percentage of total population) for each of three age groups, 0—40, 41—70, and >70. The first group could be interpreted very roughly as youth.
I think Kumar would say that the actual world would be better the closer it would resemble the Steady World, and worse the closer it would resemble the Shrinking World. Youth drains away in the Shrinking World. Although youth doesn’t decline as quickly as one might expect in terms of percentage of total population, in absolute terms it is decimated. There are also many fewer people in The Shrinking World.
The Shrinking World The Steady World
2064 (peak year)
Global population: 9,732,922,689 9,732,922,689
0–40: 4,849,520,815 (49.8%) 4,849,520,815 (49.8%)
41–70: 3,505,824,099 (36.0%) 3,505,824,099 (36.0%)
>70: 1,377,577,775 (14.2%) 1,377,577,775 (14.2%)
2100 (reference point)
Global population: 8,785,553,665 9,732,922,689
0–40: 3,766,185,119 (42.9%) 4,708,704,836 (48.4%)
41–70: 3,290,666,955 (37.5%) 3,302,279,742 (33.9%)
>70: 1,728,701,591 (19.7%) 1,720,886,824 (17.7%)
2200 (end year)
Global population: 4,75,593,134 9,732,922,689
0–40: 1,691,442,373 (35.6%) 4,932,844,441 (50.7%)
41–70: 1,678,602,241 (35.3%) 3,127,502,899 (32.1%)
>70: 1,387,548,520 (29.2%) 1,671,524,062 (17.2%)
Evaluative questions: a population ethics perspective
It is easy to imagine the Shrinking World lacking economic growth or innovation, but how might that translate into losses (or gains) in what has intrinsic value?
From a population ethics perspective, individual welfare has intrinsic value. To evaluate any population, we must ask: how are its members faring? Some lives are better than others for an individual. A life can be good, bad, or neutral for an individual. A good life is worth living; a bad life is worth not living; a neutral life is neither. There are different theories of individual welfare—hedonistic theories tie welfare to pleasure, preference-satisfaction theories to desire-fulfillment. But on any theory, there are clear cases of lives going better and worse, and of good and bad lives. A life free from extreme misery, enjoying health, love, and achievement throughout seems good. A life completely consumed throughout by unrelenting pain or severe deprivation seems bad.
Furthermore, some philosophers think that there can be both intrapersonal and interpersonal welfare comparisons on a ratio scale. This is a strong but defensible claim, though I won’t try to defend it here. (See Suppes 1966 for discussion.)
One implication of it is that we can, in principle, represent the total welfare of a population as the sum of individuals’ welfare levels. It is then possible to develop tools that approximate changes in a population’s total welfare. We could use life-satisfaction surveys that rely on subjective reporting, or Quality-Adjusted Life Years, which are commonly used to approximate the collective benefits of health interventions, for instance. Such tools are flawed, but potentially useful.
Next, we need to evaluate whole populations—sets of lives. Philosophers propose different population axiologies for this purpose. Two well-known ones are:
- The Total View: one population is better than another if and only if it has a greater total of individuals’ welfare (worse if and only if its total is less, and equally as good if and only if the totals are the same)
- The Average View: one population is better than another if and only if it has a higher average welfare (worse if and only if its average is lower, and equally as good if and only if they have the same average)
These are not the only views. Others give weight to both total and average welfare; some give additional weight to increasing the welfare of individuals at lower welfare levels, reflecting special concern for the worse off; some give weight to equality. And there are many other kinds of theories. (See Chappell, Meissner, and MacAskill, 2023, for discussion.)
The main point is that a population axiology is helpful for estimating how things like slowing economic growth or technological innovation affect the value of the Shrinking and Steady Worlds.
One reason to believe that The Shrinking World is worse
Instead of examining every way a shrinking population might affect population value, I will focus on one important factor.
Many population axiologies, not just the Total View, hold that adding more people with good lives makes a world better, other things being equal (i.e. focusing only on the presence of absence of these lives). This suggests the Shrinking World is worse than the Steady World: many fewer people are ever born in that world. I would optimistically assume that their lives would be overall good, though it is very hard to say how good.
Some population axiologies deny that additional well-off people make the world better, but these views have serious problems. The Average View is one example. On this view, adding people with good lives can make a population worse, other things being equal, if it lowers average welfare. (Importantly, Christian Tarsney and Teruji Thomas (2024) argue that when we consider the populations of whole worlds, including past individuals, most axiologies, including the Average and Total Views, may converge in their evaluations: https://journals.publishing.umich.edu/ergo/article/id/5714/.)
However, the Average View has absurd implications. Here is one illustration based on an example from the philosopher Derek Parfit (1984). Imagine a terrible world in which everyone has a bad life. Now imagine a world just like it but with twice (or three, or four times) as many miserable individuals. Average welfare in each world is the same, so the Average View says the two worlds are equally bad. That verdict is absurd. The second world, with many more hellish lives, is worse. Average welfare cannot be all that matters.
Some readers might reject The Average View but still balk at the suggestion that adding more well-off people can make a world better. They might think we can make the world better by improving individual’s lives, but adding new well-off people to the world is neutral.
But, as John Broome showed in his book Weighing Lives (2004), this idea is hard to defend. The neutrality of additional well-off people might be understood as equal goodness. Adding some person, Alice, to the world at positive welfare level L1 would be equally as good as not adding her at all. Similarly, adding Alice at some higher welfare level L2 (a better life) would be equally as good as not adding her. But ‘equally as good’ is a transitive relation. If adding Alice at L1 is equally as good as not adding her, and adding Alice at L2 is equally as good as not adding her, then adding Alice at L1 is equally as good as adding Alice at L2. That is false. Alice’s life at L2 is better than her life at L1. Surely, a world with a better off Alice is better than (not equally as good as) a world with a worse off Alice.
Neutrality might instead be understood as incommensurability: a world with a well-off Alice is neither better, worse, nor equally as good as one without her. But, as Broome shows (Weighing Lives, 164—186) this too creates problems. The argument is complicated, and I won’t rehearse it here, but the upshot is this: A world W without Alice can be incommensurable with a world W+ that includes Alice, even if W+ is considerably worse than W in other respects. Her neutral existence could offset others’ welfare losses; moreover, there may be no limit to how much welfare loss could be “swallowed up” by the neutrality of people’s existence.
A shrinking population would mean fewer good lives, and in my view, this is a real value loss. That doesn’t settle any policy questions, but it helps underscore why Kumar’s concerns matter—and why progressive pronatalism deserves consideration.
Reasons to doubt that The Shrinking World is worse
Almost every population axiology holds that the world is worse if it contains more individuals with bad lives. ‘Individuals’ includes non-human animals, many of which have welfare levels. One reason to think the Shrinking World is better concerns animal agriculture. If factory-farmed animals endure suffering, and have bad lives, then fewer humans will probably mean fewer bad animal lives.
Consumption is tied to population size and income growth. A larger, richer population likely means vastly more chickens, most confined in intensive farms. It may be that for each missing person-year in the Shrinking World, there is some missing chicken life-years. One critical question is how many? Another is whether those chicken-years’ badness outweighs the goodness of human-years. Without approximate answers, we should be uncertain that the Steady World is better.
Would pronatalist policies help reduce animal suffering? Perhaps. More people would also mean more young, politically active people. That might spur reform. But it is unclear to what extent this would offset higher demand for meat in the Steady World.
The second reason to doubt that the Shrinking World is worse concerns technological risk, including risk of human extinction. Kumar emphasizes technological innovation, which includes life-improving but also risky technologies, e.g., from synthetic biology and unaligned artificial intelligence. Kumar is no doubt correct that a world with more people would have a faster pace of both technological development and social progress. But as Hilary Greaves points out, a crucial question is whether a larger population would speed up social progress relative to technological progress, or vice versa. If the former, a larger population may be more resilient, and carry lower technological risk, including risk of human extinction; if the latter, then it may carry greater risk (https://users.ox.ac.uk/~mert2255/popular/population-priority.pdf).
It seems to me that humanity has been much better at creating new technologies than at managing their risk. Innovation seems to outpace human wisdom in using it. Perhaps a slower pace of technological progress would give us more time to plan, develop safeguards, and implement best practices to reduce risk of self-destruction.
The concern about technological risk becomes much more important when we consider longer timescales. So far, I’ve compared only two different possible narrow 136-year “slices” of the future, between 2064 and 2200. But if we consider humanity’s entire potential future, the evaluation changes. The moral stakes explode. Even on conservative estimates, the number of possible future lives—human, digital, or otherwise—dwarfs everything so far. (See, e.g., Greaves and MacAskill: https://academic.oup.com/book/60794/chapter/530063399.) If population decline (or increase) raises the risk of extinction, the expected value of preventing (or bring about) that decline could be astronomical.
But what about animal suffering? I confess that my thoughts about this are both speculative and optimistic, but for me, considering the longer timeframe changes the moral weight of this consideration. In the “short term”, fewer humans may mean fewer factory-farmed animals and less suffering. But if we look far enough ahead, that argument weakens. Technological progress—including in food production—is already making it possible to decouple nutrition from animal suffering. Moreover, if humanity advances to a stage where we are largely non-biological beings, animal agriculture may become completely obsolete. In any case, it is hard to imagine that after, say, a thousand years, factory farming would remain humanity’s dominant food system. If civilization endures, agricultural suffering is likely to diminish to negligible levels.
So on very long timescales, technological risk, and especially extinction risk, seems to be the dominant consideration. If we are uncertain how this risk varies with population size, we should be uncertain whether the effect, on the expected value of the entire future, of widespread pronatalist policies would be positive or negative. I welcome your thoughts!
References
Broome, J. (2004). Weighing lives. Oxford University Press.
Chappell, R. Y., Meissner, D., & MacAskill, W. (2023). Population Ethics. In R. Y. Chappell, D. Meissner, & W. MacAskill (Eds.), An Introduction to Utilitarianism. Retrieved from https://www.utilitarianism.net/population-ethics.
Greaves, H. (2017). Population as a global priority? Unpublished manuscript, University of Oxford. Retrieved from https://users.ox.ac.uk/~mert2255/popular/population-priority.pdf
Greaves, H., & MacAskill, W. (2021). The case for strong longtermism (GPI Working Paper No. 5-2021). Global Priorities Institute.
Parfit, D. (1984). Reasons and persons. Oxford University Press.
Suppes, P. (1966). “Some Formal Models of Interpersonal Comparison of Utility.” Behavioral Science 11(5): 306–318.
Tarsney, C., & Thomas, T. (2024). Non-Additive Axiologies in Large Worlds. Ergo: An Open Access Journal of Philosophy, 11(9). https://doi.org/10.3998/ergo.5714
United Nations, Department of Economic and Social Affairs, Population Division. (2024). World population prospects 2024. UN DESA. https://population.un.org/wpp.
Vollset, S. E., Goren, E., Yuan, C. et al. (2020). “Fertility, mortality, migration, and population scenarios for 195 countries and territories from 2017 to 2100: A forecasting analysis for the Global Burden of Disease Study.” The Lancet 396(10258): 1285–1306. https://doi.org/10.1016/S0140-6736(20)30677-2.
This is all really interesting. Thanks to everyone involved in this exchange!
I want to begin by admitting that I may be completely out of my depth here. But I also want to chime in to report that I was struck, in reading both Kumar’s article and Campbell’s post, by a shared set of assumptions that appear to be doing a lot of work in framing the issues. By making these assumptions explicit, I hope to bring out some reasons to think that the terms of this debate may be needlessly restrictive. Some reasonable possibilities remain unconsidered, and when we do consider these possibilities, we see that things may look much different than when we ignore them.
Campbell says, echoing claims made by Kumar, that “[i]t is easy to imagine the Shrinking World lacking economic growth or innovation.” The reason, as Campbell puts it, again paraphrasing Kumar, is that “a world with more people would have a faster pace of both technological development and social progress.” As Kumar puts it: “the fundamental reason that population decline will be a curse rather than a blessing [is] the loss of young people.” And the reason for this is that: “To thrive, societies need young people. New generations drive economic growth, pioneer technologies, challenge outdated moral views, create art, and advance social change. They’re more likely to take risks, embrace new ideas, and imagine different futures. When we talk about population decline, what we’re really talking about is the gradual dissipation of this vital social force.”
Perhaps they are both correct about all of this. But then again, perhaps not. It all depends, or so it seems to me, on whether or not the future is continuous with the past in crucial respects.
I already said I may be out of my depth here, and one thing I meant by that is that I have a hard time articulating exactly what I take the relevant assumptions to be here, let alone what bothers me about them. So, I’ll try and gesture at my complaint by way of invoking some comments by Toni Morrison in a famous interview with Time magazine several decades ago (https://time.com/archive/6702572/toni-morrison-the-pain-of-being-black/).
Towards the end of the interview, Morrison is asked about one of the hot topics du jour (this was in 1989): teen pregnancy. The interviewer frames this is as a “depressing” issue.
Q. And teenage pregnancies?
A. Everybody’s grandmother was a teenager when they got pregnant. Whether they were 15 or 16, they ran a house, a farm, they went to work, they raised their children.
Q. But everybody’s grandmother didn’t have the potential for living a different kind of life. These teenagers — 16, 15 — haven’t had time to find out if they have special abilities, talents. They’re babies having babies.
A. The child’s not going to hurt them. Of course, it is absolutely time consuming. But who cares about the schedule? What is this business that you have to finish school at 18? They’re not babies. We have decided that puberty extends to what — 30? When do people stop being kids? The body is ready to have babies, that’s why they are in a passion to do it. Nature wants it done then, when the body can handle it, not after 40, when the income can handle it.
Q. You don’t feel that these girls will never know whether they could have been teachers, or whatever?
A. They can be teachers. They can be brain surgeons. We have to help them become brain surgeons. That’s my job. I want to take them all in my arms and say, “Your baby is beautiful and so are you and, honey, you can do it. And when you want to be a brain surgeon, call me — I will take care of your baby.” That’s the attitude you have to have about human life. But we don’t want to pay for it.
I’m not here to litigate the correctness of Morrison’s view. What I want to bring out is that she challenges a set of assumptions that stand behind the framing of the question. To paraphrase: teenage pregnancies are only worrisome if you hold fixed certain social facts, such as that one needs to complete a certain amount of schooling uninterrupted so that one can enter the workforce, that entering the workforce at a particular age is necessary to earning the money one needs to raise a family, and that raising a family is the responsibility of the parents alone. As Morrison points out, these assumptions have not always held true, not even in the recent past. So, there is room to ask why they must hold true now.
Similarly, I wonder why we need to hold fixed the sorts of assumptions that Kumar (and to some degree Campbell) is making about what drives change and growth. Even if it’s always been true that young people have pushed society forward in these ways, must that be true in the future as well?
Morrison points out that teen bodies are ready to have babies; it’s our social arrangements that make it difficult to raise children at the age we are best suited, biologically speaking, to do so. If we imagine different social arrangements, then we can countenance the possibility that perhaps teen pregnancy wouldn’t be depressing, or even a barrier to quality of life or various good outcomes, including successful careers that both contribute to society and take a great deal of training and preparation.
Shifting now to the topic of this discussion, what happens when we consider the possibility that it’s not due to biology or anything else intrinsic to youth that makes it so that there are correlations between higher proportions of young people and progress? What happens when we take seriously the possibility that these correlations are due, at least in part, to mutable social arrangements? What makes it the case that a greater proportion of young people is correlated with greater social, economic, and technological progress? And are these factors, whatever they happen to be, immutable for all time? Or are they things that might change? That we might change? That we may have good reason to change?
My point is not that correct answers to these questions show Kumar and Campbell to be mistaken. My point is that pursuit of these questions is closed off from the get-go. Both Kumar and Campbell simply assume that the future will be like the present and recent past in certain crucial respects. Moreover, these assumptions may be doing a lot of work in the context of their arguments. In parallel with Morrison’s point about teen pregnancy, it may be the case that aging populations look so worrisome simply because we lack the imagination to consider how the social world may be different. What happens if we open our mind’s a bit wider? Perhaps we’ll reach the same conclusion. Or perhaps we’ll see a different set of solutions than those that are on the table given the way the discussion is currently framed.
Another possibility is that I’m simply missing something. I’d be just as interested in learning what I’m missing as I would be in learning what others think about the effects of taking a more expansive view of future possibilities in the context of discussing population ethics. Thanks again to everyone involved here.
Hi Tim, thanks for this excellent discussion (and to Victor Kumar for his Aeon article inspiring it).
One question I have is whether you see life-saving efforts as similarly vexed. Ordinarily, we think that reducing child mortality is straightforwardly good. But it seems like your two “reasons to doubt that The Shrinking World is worse” are equally reasons to doubt that saving lives (on present margins) is good. If we should welcome a smaller population, then it seems like each premature death would be grounds for (instrumental) celebration—we’re getting closer to the target!
That seems a pretty uncomfortable verdict. And maybe one thing that’s going on is that we ordinarily feel some moral pressure to default to an optimistic view about the value of others’ lives. (It would seem perverse to see the child drowning in Singer’s pond, and respond with a sigh of relief: “There goes one less meat-eater and/or potentially dangerous scientific innovator!”) If so, it’s an interesting question whether we should be similarly optimistic about potential lives, and thus reject the sorts of speculative instrumental reasons you’re pointing to as somehow inappropriate or inapt to outweigh the intrinsic value of the lives that could be brought about by pro-natalist policies.
On the other hand, it’s a bit puzzling how we could have moral reasons to disregard empirical hypotheses that could very well be true (and would carry high moral stakes if they were). So maybe my discomfort here is *mere* discomfort, and we really should all feel more conflicted about saving lives in general?
Hi Richard,
As usual, your questions are both fantastic and difficult. You probably know, but some readers may not, that an article on this topic by Michael Plant, “The Meat Eater Problem”, whcih appeared in the Journal of Controversial Ideas in 2022. I’m not sure if the hyperlink will appear in the chat box, but readers can find the paper here: (https://journalofcontroversialideas.org/article/2/2/206).
I think I would agree that if we should default to an optimistic view about the contributive value of others’ lives when deciding to avert deaths, then it makes sense to be similarly optimistic about the contributive value of potential lives. I don’t know if we should default to an optimistic view in either case. I WANT to be optimistic in both cases, but that’s just a psychological report.
Here’s one way that you might be led to the conclusion that your discomfort is *mere* discomfort. You mention that it would seem perverse to see the drowning child and respond with a sigh of relief “there goes one less meat-eater” (ignoring dangerous innovation for the moment). Would it still seem perverse if you somehow knew that the contribution to aggregate well-being of saving the child would be net negative, due to impact on animals? What if you didn’t know but you accepted Michael Plant’s estimate (which he takes from a 2005 paper by Matheny and Chan) that a typical human year of eating meat corresponds to about five years of factory-farmed chicken suffering, and the only info you have about the drowning child warrants treating her a typical meat eater? Next, what if you didn’t accept Plant’s estimate, but gave it some credence, and beleived that the impact of the child dying on aggregate well-being (including animals) was probably net positive? Perhaps it would seem perverse to have the thought you mention in each of these cases. But then we should wonder whether what makes the thought seem perverse is something other than the expected morally significant consequences of the child drowning. Of course someone (not me, and I guess not you) might argue that that thought is not what the “virtuous person” would think, irrespective of beliefs about the further consequences on aggregate well-being.
Here are a few other thoughts. These don’t really provide adequte answers to your challenge, but may be useful for some readers.
First, there is at least logical space for the view that we should avert deaths (and be happy when they are averted) despite the meat eater and techological risk worries, and yet try to bring about (and be happy about) a smaller population in which fewer people are ever born, given those same worries. One could, for instance, think that there is stronger moral reason to extend good lives than create new ones, other things being equal. I wouldn’t want to commit to this position, but I suspect some would. Alternatively, one could plausibly think that, generally speaking, one more death is worse than one fewer person ever being born, due to the negative externalities of death. So it is possible (but probably really hard) to do the cost-benefit analysis in such a way that one can feel good about averting deaths, but either bad, or conflicted, about more people being born.
Second, suppose you took seriously both Plant’s meat eater worry and the worry about technological risk. I suppose it would still be possible to feel good about averting deaths in some parts of the world. Perhaps saving the lives of the world’s poorest people would contribute hardly anything to either problem, and so the expected value (even for the whole world) of saving one of the world’s poorest would be positive.
Finally, the most attractive option here would be to find a good justification for the optimistic take about saving lives (and perhaps also creating more lives), even given the meat eater and technological risk problems. I guess if one could justify the claim that a larger population has higher expected value for the whole world than a smaller one, then one could avoid the discomfort altogether. I’m holding out for this.
Hi Ben,
Thanks for your really toughtful post.
I was mostly taking Kumar’s claims about the connection between youth and innovation and social progress for granted. I do agree with those claims. But I confess that I don’t know what the fundamental explanation of the connection is. I can imagine different possibilities. One is that the connection is mostly biological. According to this (possible) explanation, youth is intrinsically linked to novelty-seeking, creativity, risk-taking, and plasticity, and these traits have neurobiological underpinnings—greater dopaminergic activity, faster learning rates, more cognitive flexibility, etc. Another is that the social roles that young people occupy in society, such as being less invested in status quo, having less institutional power, and having more freedom to experiment make them situationally (but not necessarily neurobiologically) more innovative. Finally, I can imagine a hybrid explanation that points to both biological and social factors. On either the second or third explanations, there would be a possibility to change the relationship between age and innovation.
I can imagine social arrangements that seek more to reward experimentation, for instance, and seek less to punish failure. I can imagine a society where lifelong education and social permeability are emphasized much more, and perhaps heavily subsidized, so that workers do not need to entrench in very specialized positions in mid-life, which seems to be the norm in most societies.
Finally, we might consider technological changes. If machine learning systems can extend memory and reasoning, older individuals might continue to innovate well beyond the age when working memory or fluid intelligence with otherwise decline. We can even imagine futures where machines, and not humans do most of this work, although that raises a bunch of further issues.
Basically, you are right that there are some important questions that Kumar (and me) haven’t raised in this discussion. Generally speaking, it is hard to hypothesize about the future, especially over long timescales. We are all out of our depth here. In responding to Kumar’s article, I therefore had to choose which assumptions and claims to question, and which to leave untouched. You’re raising a good point about one that I left untouched. I don’t think I know enough about the issue to say more about it, though.
Thanks for this thoughtful reply, Tim. You’re absolutely correct that we have to choose which starting points to accept and which to question. And it seems to me that you (and Kumar) chose in line with what most folks accept. I suppose I was trying to think through some of the implications of making different choices. Your helpful comments about different possible explanations of the connection between youth and progress are very interesting in connection with this. Thanks!