In What We Owe the Future, William MacAskill delves into population ethics in a chapter titled “Is It Good to Make Happy People?” (Chapter 8). As he writes at the outset of the chapter, our views on population ethics matter greatly for our priorities, and hence it is important that we reflect on the key questions of population ethics. Yet it seems to me that the book skips over some of the most fundamental and most action-guiding of these questions. In particular, the book does not broach questions concerning whether any purported goods can outweigh extreme suffering — and, more generally, whether happy lives can outweigh miserable lives — even as these questions are all-important for our priorities.
The Asymmetry in population ethics
A prominent position that gets a very short treatment in the book is the Asymmetry in population ethics (roughly: bringing a miserable life into the world has negative value while bringing a happy life into the world does not have positive value — except potentially through its instrumental effects and positive roles).
The following is, as far as I can tell, the main argument that MacAskill makes against the Asymmetry (p. 172):
If we think it’s bad to bring into existence a life of suffering, why should we not think that it’s good to bring into existence a flourishing life? I think any argument for the first claim would also be a good argument for the second.
This claim about “any argument” seems unduly strong and general. Specifically, there are many arguments that support the intrinsic badness of bringing a miserable life into existence that do not support any intrinsic goodness of bringing a flourishing life into existence. Indeed, many arguments support the former while positively denying the latter.
One such argument is that the presence of suffering is bad and morally worth preventing while the absence of pleasure is not bad and not a problem, and hence not morally worth “fixing” in a symmetric way (provided that no existing beings are deprived of that pleasure).
A related class of arguments in favor of an asymmetry in population ethics is based on theories of wellbeing that understand happiness as the absence of cravings, preference frustrations, or other bothersome features. According to such views, states of untroubled contentment are just as good — and perhaps even better than — states of intense pleasure.
These views of wellbeing likewise support the badness of creating miserable lives, yet they do not support any supposed goodness of creating happy lives. On these views, intrinsically positive lives do not exist, although relationally positive lives do.
Another point that MacAskill raises against the Asymmetry is an example of happy children who already exist, about which he writes (p. 172):
if I imagine this happiness continuing into their futures—if I imagine they each live a rewarding life, full of love and accomplishment—and ask myself, “Is the world at least a little better because of their existence, even ignoring their effects on others?” it becomes quite intuitive to me that the answer is yes.
However, there is a potential ambiguity in this example. The term “existence” may here be understood to either mean “de novo existence” or “continued existence”, and interpreting it as the latter is made more tempting by the fact that 1) we are talking about already existing beings, and 2) the example mentions their happiness “continuing into their futures”.
This is relevant because many proponents of the Asymmetry argue that there is an important distinction between the potential value of continued existence (or the badness of discontinued existence) versus the potential value of bringing a new life into existence.
Thus, many views that support the Asymmetry will agree that the happiness of these children “continuing into their futures” makes the world better, or less bad, than it otherwise would be (compared to a world in which their existing interests and preferences are thwarted). But these views still imply that the de novo creation (and eventual satisfaction) of these interests and preferences does not make the world better than it otherwise would be, had they not been created in the first place. (Some sources that discuss or defend these views include Singer, 1980; Benatar, 1997; 2006; Fehige, 1998; Anonymous, 2015; St. Jules, 2019; Frick, 2020.)
A proponent of the Asymmetry may therefore argue that the example above carries little force against the Asymmetry, as opposed to merely supporting the badness of preference frustrations and other deprivations for already existing beings.
Questions about outweighing
Even if one thinks that it is good to create more happiness and new happy lives all else equal, this still leaves open the question as to whether happiness and happy lives can outweigh suffering and miserable lives, let alone extreme suffering and extremely bad lives. After all, one may think that more happiness is good while still maintaining that happiness cannot outweigh intense suffering or very bad lives — or even that it cannot outweigh the worst elements found in relatively good lives. In other words, one may hold that the value of happiness and the disvalue of suffering are in some sense orthogonal (cf. Wolf, 1996; 1997; 2004).
As mentioned above, these questions regarding tradeoffs and outweighing are not raised in MacAskill’s discussion of population ethics, despite their supreme practical significance. One way to appreciate this practical significance is by considering a future in which a relatively small — yet in absolute terms vast — minority of beings live lives of extreme and unrelenting suffering. This scenario raises what I have elsewhere (sec. 14.3) called the “Astronomical Atrocity Problem”: can the extreme and incessant suffering of, say, trillions of beings be outweighed by any amount of purported goods? (See also this short excerpt from Vinding, 2018.)
After all, an extremely large future civilization would contain such (in absolute terms) vast amounts of extreme suffering in expectation, which renders this problem frightfully relevant for our priorities.
MacAskill’s chapter does discuss the Repugnant Conclusion at some length, yet the Repugnant Conclusion does not explicitly involve any tradeoffs between happiness and suffering, and hence it has limited relevance compared to, for example, the Very Repugnant Conclusion (roughly: that arbitrarily many hellish lives can be “compensated for” by a sufficiently vast number of lives that are “barely worth living”).
Indeed, the Very Repugnant Conclusion and similar such “offsetting conclusions” would seem more relevant to discuss both because 1) they do explicitly involve tradeoffs between happiness and suffering, or between happy lives and miserable lives, and because 2) MacAskill himself has stated that he considers the Very Repugnant Conclusion to be the strongest objection against his favored view, and stronger objections generally seem more worth discussing than do weaker ones.
Popular support for significant asymmetries in population ethics
MacAskill briefly summarizes a study that surveyed people’s views on population ethics. Among other things, he writes the following about the findings of the study (p. 173):
these judgments [about the respective value of creating happy lives and unhappy lives] were symmetrical: the experimental subjects were just as positive about the idea of bringing into existence a new happy person as they were negative about the idea of bringing into existence a new unhappy person.
While this summary seems accurate if we only focus on people’s responses to one specific question in the survey (cf. Caviola et al., 2022, p. 9), there are nevertheless many findings in the study that suggest that people generally do endorse significant asymmetries in population ethics.
Specifically, the study found that people on average believed that considerably more happiness than suffering is needed to render a population or an individual life worthwhile, even when the happiness and suffering were said to be equally intense (Caviola et al., 2022, p. 8). The study likewise found that participants on average believed that the ratio of happy to unhappy people in a population must be at least 3-to-1 for its existence to be better than its non-existence (Caviola et al., 2022, p. 5).
Another relevant finding is that people generally have a significantly stronger preference for smaller over larger unhappy populations than they do for larger over smaller happy populations, and the magnitude of this difference becomes greater as the populations under consideration become larger (Caviola et al., 2022, pp. 12-13).
In other words, people’s preference for smaller unhappy populations becomes stronger as population size increases, whereas the preference for larger happy populations becomes less strong as population size increases, in effect creating a strong asymmetry in cases involving large populations (e.g. above one billion individuals). This finding seems particularly relevant when discussing laypeople’s views of population ethics in a context that is primarily concerned with the value of potentially vast future populations.
Moreover, a pilot study conducted by the same researchers suggested that the framing of the question plays a major role for people’s intuitions (Caviola et al., 2022, “Supplementary Materials”). In particular, the pilot study (n=172) asked people the following question:
Suppose you could push a button that created a new world with X people who are generally happy and 10 people who generally suffer. How high would X have to be for you to push the button?
When the question was framed in these terms, i.e. in terms of creating a new world, people’s intuitions were radically more asymmetric, as the median ratio then jumped to 100-to-1 happy to unhappy people, which is a rather pronounced asymmetry.
In sum, it seems that the study that MacAskill cites above, when taken as a whole, mostly finds that people on average do endorse significant asymmetries in population ethics. I think this documented level of support for asymmetries would have been worth mentioning.
(Other surveys that suggest that people on average affirm a considerable asymmetry in the value of happiness vs. suffering and good vs. bad lives include the Future of Life Institute’s Superintelligence survey (n=14,866) and Tomasik, 2015 (n=99).)
The discussion of moral uncertainty excludes asymmetric views
Toward the end of the chapter, MacAskill briefly turns to moral uncertainty, and he ends his discussion of the subject on the following note (p. 187):
My colleagues Toby Ord and Hilary Greaves have found that this approach to reasoning under moral uncertainty can be extended to a range of theories of population ethics, including those that try to capture the intuition of neutrality. When you are uncertain about all of these theories, you still end up with a low but positive critical level [of wellbeing above which it is a net benefit for a new being to be created for their own sake].
Yet the analysis in question appears to wholly ignore asymmetric views in population ethics. If one gives significant weight to asymmetric views — not to mention stronger minimalist views in population ethics — the conclusion of the moral uncertainty framework is likely to change substantially, perhaps so much so that the creation of new lives is generally not a benefit for the created beings themselves (although it could still be a net benefit for others and for the world as a whole, given the positive roles of those new lives).
Similarly, even if the creation of unusually happy lives would be regarded as a benefit from a moral uncertainty perspective that gives considerable weight to asymmetric views, this benefit may still not be sufficient to counterbalance extremely bad lives, which are granted unique weight by many plausible axiological and moral views (cf. Mayerfeld, 1999, pp. 114-116; Vinding, 2020, ch. 6).
Ajantaival, T. (2021/2022). Minimalist axiologies. Ungated
Anonymous. (2015). Negative Utilitarianism FAQ. Ungated
Benatar, D. (1997). Why It Is Better Never to Come into Existence. American Philosophical Quarterly, 34(3), pp. 345-355. Ungated
Benatar, D. (2006). Better Never to Have Been: The Harm of Coming into Existence. Oxford University Press.
Contestabile, B. (2022). Is There a Prevalence of Suffering? An Empirical Study on the Human Condition. Ungated
DiGiovanni, A. (2021). A longtermist critique of “The expected value of extinction risk reduction is positive”. Ungated
Fehige, C. (1998). A pareto principle for possible people. In Fehige, C. & Wessels U. (eds.), Preferences. Walter de Gruyter. Ungated
Frick, J. (2020). Conditional Reasons and the Procreation Asymmetry. Philosophical Perspectives, 34(1), pp. 53-87. Ungated
Future of Life Institute. (2017). Superintelligence survey. Ungated
Gloor, L. (2016). The Case for Suffering-Focused Ethics. Ungated
Gloor, L. (2017). Tranquilism. Ungated
Hurka, T. (1983). Value and Population Size. Ethics, 93, pp. 496-507.
James, W. (1901). Letter on happiness to Miss Frances R. Morse. In Letters of William James, Vol. 2 (1920). Atlantic Monthly Press.
Knutsson, S. (2019). Epicurean ideas about pleasure, pain, good and bad. Ungated
MacAskill, W. (2022). What We Owe The Future. Basic Books.
Mayerfeld, J. (1999). Suffering and Moral Responsibility. Oxford University Press.
Parfit, D. (1984). Reasons and Persons. Oxford University Press.
Sherman, T. (2017). Epicureanism: An Ancient Guide to Modern Wellbeing. MPhil dissertation, University of Exeter. Ungated
Singer, P. (1980). Right to Life? Ungated
St. Jules, M. (2019). Defending the Procreation Asymmetry with Conditional Interests. Ungated
Tomasik, B. (2015). A Small Mechanical Turk Survey on Ethics and Animal Welfare. Ungated
Tsouna, V. (2020). Hedonism. In Mitsis, P. (ed.), Oxford Handbook of Epicurus and Epicureanism. Oxford University Press.
Vinding, M. (2018). Effective Altruism: How Can We Best Help Others? Ratio Ethica. Ungated
Vinding, M. (2020). Suffering-Focused Ethics: Defense and Implications. Ratio Ethica. Ungated
Wolf, C. (1996). Social Choice and Normative Population Theory: A Person Affecting Solution to Parfit’s Mere Addition Paradox. Philosophical Studies, 81, pp. 263-282.
Wolf, C. (1997). Person-Affecting Utilitarianism and Population Policy. In Heller, J. & Fotion, N. (eds.), Contingent Future Persons. Kluwer Academic Publishers. Ungated
Wolf, C. (2004). O Repugnance, Where Is Thy Sting? In Tännsjö, T. & Ryberg, J. (eds.), The Repugnant Conclusion. Kluwer Academic Publishers. Ungated
Further arguments against a moral symmetry between happiness and suffering are found in Mayerfeld, 1999, ch. 6; Vinding, 2020, sec. 1.4 & ch. 3.
On some views of wellbeing, especially those associated with Epicurus, the complete absence of any bothersome or unpleasant features is regarded as the highest pleasure, Sherman, 2017, p. 103; Tsouna, 2020, p. 175. Psychologist William James also expressed this view, James, 1901.
I am not saying that the “continued existence” interpretation is necessarily the most obvious one to make, but merely that there is significant ambiguity here that is likely to confuse many readers as to what is being claimed.
Moreover, a proponent of minimalist axiologies may argue that the assumption of “ignoring all effects on others” is so radical that our intuitions are unlikely to fully ignore all such instrumental effects even when we try to, and hence we may be inclined to confuse 1) the relational value of creating a life with 2) the (purported) intrinsic positive value contained within that life in isolation — especially since the example involves a life that is “full of love and accomplishment”, which might intuitively evoke many effects on others, despite the instruction to ignore such effects.
Chapter 9 in MacAskill’s book does review some psychological studies on intrapersonal tradeoffs and preferences (see e.g. p. 198), but these self-reported intrapersonal tradeoffs do not necessarily say much about which interpersonal tradeoffs we should consider plausible or valid. Nor do these intrapersonal tradeoffs generally appear to include cases of extreme suffering, let alone an entire lifetime of torment (as experienced, for instance, by many of the non-human animals whom MacAskill describes in Chapter 9). Hence, that people are willing to make intrapersonal tradeoffs between everyday experiences that are more or less enjoyable says little about whether some people’s enjoyment can morally outweigh the intense suffering or extremely bad lives endured by others. (In terms of people’s self-reported willingness to experience extreme suffering in order to gain happiness, a small survey (n=99) found that around 45 percent of respondents would not experience even a single minute of extreme suffering for any amount of happiness; and that was just the intrapersonal case — such suffering-for-happiness trades are usually considered less plausible and less permissible in the interpersonal case, cf. Mayerfeld, 1999, pp. 131-133; Vinding, 2020, sec. 3.2.)
Individual ratings of life satisfaction are similarly limited in terms of what they say about intrapersonal tradeoffs. Indeed, even a high rating of momentary life satisfaction does not imply that the evaluator’s life itself has overall been worth living, even by the evaluator’s own standards. After all, one may report a very high quality of life yet still think that the good part of one’s life cannot outweigh one’s past suffering. It is thus rather limited what we can conclude about the value of individual lives, much less the world as a whole, based on people’s momentary ratings of life satisfaction.
Finally, MacAskill also mentions various improvements that have occurred in recent centuries as a reason to be optimistic about the future of humanity in moral and evaluative terms. Yet it is unclear whether any of the improvements he mentions involve genuine positive goods, as opposed to representing a reduction of bads, e.g. child mortality, poverty, totalitarian rule, and human slavery (cf. Vinding, 2020, sec. 8.6).
Some formulations of the Repugnant Conclusion do involve tradeoffs between happiness and suffering, and the conclusion indeed appears much more repugnant in those versions of the thought experiment.
One might object that the Very Repugnant Conclusion has limited practical significance because it represents an unlikely scenario. But the same could be said about the Repugnant Conclusion (especially in its suffering-free variant). I do not claim that the Very Repugnant Conclusion is the most realistic case to consider. When I claim that it is more practically relevant than the Repugnant Conclusion, it is simply because it does explicitly involve tradeoffs between happiness and (extreme) suffering, which we know will also be true of our decisions pertaining to the future.
Some philosophers have explored, and to some degree supported, similar views. For example, Derek Parfit wrote (Parfit, 1984, p. 406): “When we consider the badness of suffering, we should claim that this badness has no upper limit. It is always bad if an extra person has to endure extreme agony. And this is always just as bad, however many others have similar lives. The badness of extra suffering never declines.” In contrast, Parfit seemed to consider it more plausible that the addition of happiness adds diminishing marginal value to the world, even though he ultimately rejected that view because he thought it had implausible implications, Parfit, 1984, pp. 406-412. See also Hurka, 1983; Gloor, 2016, sec. IV; Vinding, 2020, sec. 6.2. Such views imply that it is of chief importance to avoid very bad outcomes on a very large scale, whereas it is relatively less important to create a very large utopia.
This framing effect could be taken to suggest that people often fail to fully respect the radical “other things being equal” assumption when considering the addition of lives in our world. That is, people might not truly have thought about the value of new lives in total isolation when those lives were to be added to the world we inhabit, whereas they might have come closer to that ideal when they considered the question in the context of creating a new, wholly self-contained world. (Other potential explanations of these differences are reviewed in Contestabile, 2022, sec. 4; Caviola et al., 2022, “Supplementary Materials”, pp. 7-8.)
Or at least not sufficient to counterbalance the substantial number of very bad lives that the future contains in expectation, cf. the Astronomical Atrocity Problem mentioned above.
Further discussion of moral uncertainty from a perspective that takes asymmetric views into account is found in DiGiovanni, 2021.