It looks like "Conditional Reasons and the Procreation Asymmetry" was published in 2020, so it's even more recent than the other articles you mention (you probably read a pre-publication draft).
Thanks for putting this up here. One major and three minor comments.
First, and probably most importantly, I don't see how this line of reasoning gets an asymmetry. If I understand it correctly, the idea is that people need to actually exist to have interests, so if people do or will exist, we can say existence will be good/bad for them. But that gets you to actualism, it seems, but not an asymmetry. If X would have a bad life, were X to exist, I take it we shouldn't create X. But then why, if X were to have a good life, were X to exist, do we not to have reason to create X? You say you're 'agnostic' about whether those who would have good lives have an interest in existing, but I don't think you give a reason for this agnosticism, which would be the crucial thing to do.
Second, I didn't really understand the explication of Meacham's view - you said it 'solves' a cavalcade of issues on pop ethics but didn't spell out how it actually solves them. I'm also not sure if your view is different from Meacham's and, if so, how.
Third, it would be useful if you could spell out what you take (some of) the practical implications of your view to be.
Fourth, because you get stuck into the deep end quite quickly, I wonder if you should add a note that this is a relatively more 'advanced' forum post.
If X would have a bad life, were X to exist, I take it we shouldn't create X.
Yes, because in the outcome where X exists, they would have a claim to not exist (by my assumption), so this would have been a mistake, and you would want to undo the choice, if you could. Therefore, it's not a rational choice to make.
But then why, if X were to have a good life, were X to exist, do we not to have reason to create X?
Not from a situation in which X does not and will not exist, as a person-affecting rule (my claim), since X would not exist to be able to have such a claim. So, if you chose for X not to exist, then you would have no reason to change your mind, since X will not be around to give you such a reason. If you did choose for X to exist (whether or not they exist yet), this would give you a reason to not change your mind, but this reason does not apply if you'd have already chosen otherwise, so your choice to not bring X into existence would be stable; you wouldn't change your mind. See also my discussion with Stijn.
You say you're 'agnostic' about whether those who would have good lives have an interest in existing, but I don't think you give a reason for this agnosticism, which would be the crucial thing to do.
This is just my personal view, since I'm uncertain about the matter, and the argument goes through either way. That those who would have a good life have an interest in existing is compatible with some of the example theories I gave, including even consequentialist theories that always assign a negative value to new individuals, all else equal.
Second, I didn't really understand the explication of Meacham's view - you said it 'solves' a cavalcade of issues on pop ethics but didn't spell out how it actually solves them. I'm also not sure if your view is different from Meacham's and, if so, how.
I didn't plan to explain all of that, since it was just a brief overview and a reason for readers to check out the paper themselves.
Meacham's system is a full utilitarian calculus, and so can be taken to be a full ethical system. What I present here is just an argument that could defend or motivate parts of the design choices in his system, e.g. that in an outcome in which an individual doesn't exist, the harm to them is 0.
Third, it would be useful if you could spell out what you take (some of) the practical implications of your view to be.
That's a good point. There's a lot that will depend on what other views you hold, but I think broadly, the views would be more present-focused with a narrow person-affecting view, and generally more suffering-focused (including future suffering) with a wide or narrow person-affecting view. They would probably reject the astronomical waste argument, so human extinction or slowed population growth could only look bad instrumentally, and may even look good, although this depends on empirical views about whether our presence would be useful for others.
If we use a wide person-affecting view and also accept empty individualism (and the independence of irrelevant alternatives), then our comparison of outcomes is basically strong negative utilitarianism, although it need not aggregate through a sum.
If instead we take a narrow person-affecting view, then there's still reason to prevent bad lives in the future, but we can't make future individuals better off otherwise, since they'd be different individuals. So, we might focus on individuals alive today, as well as future bad lives only to prevent their existence, since we wouldn't be able to make future bad lives better otherwise.
If, on top of the narrow view, we also accept empty individualism, then we can't make individuals who exist now better off either, and the only good we can do for anyone is prevent them from existing. We would focus only on preventing bad individual person-moments.
I'd need to think more about whether there are further important practical implications beyond just those implied by the Procreation Asymmetry.
Fourth, because you get stuck into the deep end quite quickly, I wonder if you should add a note that this is a relatively more 'advanced' forum post.
I was actually hoping it would be fairly accessible, because I think a lot of EAs might be committed to a totalist view without having thought much about arguments for the alternatives, even if they might have found the Procreation Asymmetry intuitive before.
It seems that with the formulation of the Comparative Interest principle, you already assume an asymmetry. Consider the symmetric (equally reasonable) formulation, by writing ‘better’ instead of ‘worse’ and switching X and Y: An outcome X is in one way better than an outcome Y if, conditional on X, the individuals in X would have a stronger overall interest in outcome X than in Y and, conditional on Y, the individuals in Y would not have an even stronger overall interest in Y than in X.
With this formulation, the procreation asymmetry illustriation looks different: there is an arrow from non-existence to positive existence (top arrow from right to left), but no arrow from negative existence to non-existence.
Your formulation of the comparative interest principle, means that you focus on the tails of the arrows in the figure: an arrow can only be drawn if someone exists (and has interests) at the position of the tail of the arrow. My formulation focuses on the arrowheads: an arrow can only be drawn if someone exists (and has interests) at the position of the head of the arrow. There is a symmetry in choosing heads or tails, so your comparative interest principle is not suitable for a good defense of the procreation asymmetry.
I have another defense, based on my theory of variable critical level utilitarianism (https://stijnbruers.wordpress.com/2018/02/24/variable-critical-level-utilitarianism-as-the-solution-to-population-ethics/). This is a critical level utilitarianism, where now everyone is free to choose their own critical level. The condition is: everyone should be willing to accept a life at the chosen critical level. This means that no-one will choose a negative critical level. Critical levels always have to be positive. That introduces an asymmetry between the positive and the negative, and this asymmetry is at the root of the procreation asymmetry.
It seems that with the formulation of the Comparative Interest principle, you already assume an asymmetry. Consider the symmetric (equally reasonable) formulation, by writing ‘better’ instead of ‘worse’ and switching X and Y
This is a good point, and I should have put more thought into this. I think there's a pretty good reason to accept my original principle that does not apply to the modified one: mine implies a kind of stability by focusing on arrow tails, while the modified one does not seem to. I did write "stable (in a sense somewhat similar to a decision/game-theoretic one)", but didn't expand further on this or consider your modified principle. I'll do that here.
We first consider the interests of those existing in the given outcome for person-affecting reasons and then only consider the interests in the opposite direction from the other outcome as a potential defeater if the first interests actually pointed towards the other outcome. This is to ensure we don't change our minds back and forth between the two outcomes. (I see now that we might want to extend the consideration of the interests in the opposite direction to cycles of length > 2.)
That is, if X is worse than Y in my way, and you choose X, you would realize it was a mistake after considering the interests in X that you actually observe and you would wish for Y to have happened instead, and may even try to make the future closer to Y, undoing the work you did for X. In my view, it's absurd to choose outcomes which you know you will prefer to not have happened. I think we can defend this on grounds of rationality, e.g. avoiding things like money pumps and Dutch books.
On the other hand, with your modified principle, if X is better than Y, and Y happens, the interests of the individuals in X are not the same as the interests in Y, which is the outcome that actually happened. Supposing you choose Y, if there's a pull from Y to X that would cause you to change your mind about choosing Y, I claim now that it should be from the individuals in Y as they are in Y because that's your reality (the individuals in X may not even exist), but the modified principle considers overall interests from Y only in the opposite direction (for Y > X), so it has nothing to say about this.
For example (to illustrate or defend the Procreation Asymmetry directly, but I should probably not defend Comparative Interests in this way to defend the Procreation Asymmetry), if people decide to have a child they know will be forever miserable because they don't count the harm ahead of time, once the child is born (or the decision to have the child is made), the parent(s) may decide to euthanize (abort, etc.) them for the child's sake. And then, they could do this again and again and again, knowing they'll change their minds at each point, because at each point, although they might recognize the harm, they don't count it until after the decision is made. This is basically a money pump for the parents, and the modified principle allows this, while mine does not.
On the other hand, if people decide to not have a child instead of a happy one, there's no child in that outcome whose interests would push them to have the child. The child with interests in existing is in an outcome that didn't happen. The child won't be around to tell them they messed up by not bringing them into existence. So, there's no money pump.
I have another defense, based on my theory of variable critical level utilitarianism (https://stijnbruers.wordpress.com/2018/02/24/variable-critical-level-utilitarianism-as-the-solution-to-population-ethics/). This is a critical level utilitarianism, where now everyone is free to choose their own critical level. The condition is: everyone should be willing to accept a life at the chosen critical level. This means that no-one will choose a negative critical level. Critical levels always have to be positive. That introduces an asymmetry between the positive and the negative, and this asymmetry is at the root of the procreation asymmetry.
If people are choosing their own critical levels, shouldn't we just redefine negative existence and positive existence in terms of the situations/identities that correspond to those critical levels? I.e. we didn't actually put 0 at the right place in the first place, and if we did so, there would be no need to subtract a critical level. I suppose if we want to define negative and positive existence in terms of hedonistic utility (or some other values), but define critical level preferentially, this would make sense.
I'm still not perfectly convinced: there still seems to be a symmetric formulation. You describe it in terms of pushing instead of pulling. But what about the symmetry between expressions "an existing individual in X pushes the situation from X to Y", versus "an existing individual in Y pulls the situation from X to Y"? Why would there be no money pump in pulling cases if there could be a money pump in a pushing case?
That being said, my gut feeling tells me that your reference to game theoretic instability or money pumps is similar (analogous or perhaps exactly the same?) as my reference to dynamic inconsistency (subgame imperfect situations) that I described in my variable critical level utilitarianism draft paper https://stijnbruers.files.wordpress.com/2018/02/variable-critical-level-utilitarianism-1.pdf. So in the end you could be pointing at a valid argument indeed.
I'm still not perfectly convinced: there still seems to be a symmetric formulation. You describe it in terms of pushing instead of pulling. But what about the symmetry between expressions "an existing individual in X pushes the situation from X to Y", versus "an existing individual in Y pulls the situation from X to Y"? Why would there be no money pump in pulling cases if there could be a money pump in a pushing case?
In the pushing case, if you make one choice, you're sometimes compelled to change your mind: if you've chosen X, the individuals in X can push you towards Y. In the pulling case, the reasons to change your mind don't apply in the option you've chosen: if you've chosen X, the individuals in Y can't pull you towards Y, because those claims don't come from X. The claims in Y for Y over X only make a difference if there are also claims in Y for X (or something else) over Y that they defeat, which is captured by "in Y, a stronger overall interest in Y than in X" or "in Y, a stronger overall interest in X than in Y" (or equality).
I'll break it down into cases to illustrate:
Suppose "an existing individual in X pushes the situation from X to Y". Then:
1. If you choose X, an individual who exists in X has a claim to Y over X, so you have reason to change your mind to Y, and this reason only applies in X, which you've chosen. That's a reason to change your mind to Y, although it may ultimately be outweighed if there are other claims (but first by other claims in X, and then if there's an overall claim in X to Y over X, i.e. in the same direction, also by claims in Y). If there are no other reasons to be concerned with, then this is not a stable solution, since you have an overall reason in X to change your mind to Y, and no other reasons in any other outcome to change your mind.
2. If you choose Y, we don't have enough information to say anything (we don't know if the individuals in Y have claims to anything else). The claim in X to Y over X does not apply here, since you didn't choose X, although there could be other claims. If there are no other reasons to be concerned with, then this is a stable solution, since you have no reason in Y to change your mind.
Now, instead suppose "an existing individual in Y pulls the situation from X to Y". Then:
1. If you choose Y, an individual who exists in Y has a claim to Y over X, so you have a reason to not change your mind to X, but this reason only applies in Y, which you've chosen. That's a reason to not change your mind to X, although it may ultimately be outweighed if there are other claims (but first by other claims in Y, and then if there's an overall claim in Y to X over Y, i.e. in the opposite direction, also by claims in X). If there are no other reasons to be concerned with, then this is a stable solution, since you have no reason in Y to change your mind.
2. If you choose X, we don't have enough information to say anything (we don't know if the individuals in Y have claims to anything else). The claim in Y to Y over X does not apply here, since you didn't choose Y, although there could be other claims. If there are no other reasons to be concerned with, then this is a stable solution, since you have no reason in X to change your mind.
Correct me if I am wrong, but it seems like the notion of existence you are using here is either sharp or would be sharp if you had sufficient time to reflect on and refine a formalization (e.g. it formalization might be of the form "the set of physical systems A count as existing humans and everything else doesn't"). For example, your valence/non-existence diagram has a sharp line.
I'm asking this because I think the main reason why I don't hold person affecting views is that sharpness seems like it isn't a necessary assumption about my morality* and that it is more complex than at least one alternative: view all matter as potentially mattering, because, for example, there is always potential to shape the matter like a human living a good life. Viewing personhood like this also seems more accurate/practical to me (e.g. it feels like being more of a physicalist about personhood in this way would let you make more progress on questions related to digital sentience, though this claim is unsubstantiated).
Do you have any thoughts on how to weigh between "choose the morality that uses simpler-seeming assumptions" v.s. "choose the morality that matches my immediate feelings about what it means to improve the world?" I'm unsure, though I currently lean toward the former.
*maybe this is where you disagree since not having sharpness means that reproducing in the future when life isn't sucky can be good
I think I had something basically sharp in mind when I wrote this, but I don't think sharpness is essential. Rocks have no interests at all or only extremely weak ones that are very easy to outweigh. From the point of view of the rock as a rock, becoming a happy human doesn't necessarily seem better at all even assuming a symmetric theory of welfare. It could depend on exactly how you think about "interests", though, but I wouldn't use any account on which rocks have non-negligible interests as long as they remain rocks, since they care at most negligibly about anything.
We can apply Comparative Interests:
Comparative Interests: An outcome X is in one way worse than an outcome Y if, conditional on X, the individuals in X would have a stronger overall interest in outcome Y than in X and, conditional on Y, the individuals in Y would not have an equally or even stronger overall interest in X than in Y.
Suppose in X, the rock remains a rock, and in Y, the rock becomes a happy human. In X, the rock cares at most negligibly about anything, so I would say Y is at most negligibly better for the rock than X (although I'm not sure it is better at all), and this difference is small enough to be so easily outweighed that it can be ignored in practice. So, whether or not X is worse than Y basically doesn't depend on any interests the rock might have, because of how weak they'd be.
Do you have any thoughts on how to weigh between "choose the morality that uses simpler-seeming assumptions" v.s. "choose the morality that matches my immediate feelings about what it means to improve the world?" I'm unsure, though I currently lean toward the former.
I don't think there's really any good, non-arbitrary, principled and objective way to weigh these things. I guess I decide by cases on whatever just feels more right to me (although am often uncertain about that) and I try to at least avoid inconsistency. You could try to approximate something like a Kolmogorov complexity prior, but even then, you'd have a bunch of arbitrary choices to make.
not coming to exist at all would be strictly worse than coming to exist with non-maximal utility
The transitivity argument you presented shows that it's strictly better.
Nitpicks aside, thank you for sharing these ideas! I think identifying that interests (or desires associated with experiences) are the morally relevant objects rather than persons is crucial.
I think you can still treat persons as morally relevant, on top of their interests. In particular, you could think that we should weight interests within a person differently from how we weight them across persons, so that personal and interpersonal trade-offs can be treated differently. The principle of Comparative Interests I put forward doesn't make any claims about how interests should be weighted.
If you accept empty individualism, then you might just respond that each interest (or preference or experience, etc.) should be its own person, so that all trade-offs are interpersonal.
The Procreation Asymmetry consists of these two claims together:
it’s bad to bring into existence an individual who would have a bad existence, other things being equal, or the fact that an individual would have a bad existence is a reason to not bring them into existence; and
it’s at best indifferent to bring into existence an individual who would have a good existence, other things being equal, or the fact that an individual would have a good existence is not a reason to bring them into existence.
If existence can be an existential harm (according to 1. in the asymmetry), why can’t it also be an existential benefit? I.e. why should we accept 2.? If it’s worse for people who regret being born (or whose suffering outweighs their happiness) to actually be born, isn’t it better for people who are grateful for being born (or whose happiness outweighs their suffering) to actually be born? These are two immediate replies to the procreation asymmetry, and providing a satisfactory response which does not rely solely on intuition (from hypotheticals) would better make the case for the asymmetry.
I think it’s very likely that what I say here isn’t at all original; it’s probably come up often in defence of the procreation asymmetry. However, this thinking has been somewhat useful to me in defending the procreation asymmetry without just accepting it from intuition. It's roughly based on Johann Frick’s defense of the Procreation Asymmetry:
There is indeed a sort of asymmetry in the situations that can be used to defend the Procreation Asymmetry, without just referring to examples. I illustrate it here:
Consider the above graph, and suppose an individual can exist with a positive existence, exist with a neutral existence, exist with a negative existence or not exist at all. In the graph, the arrows represent when the individual in the given state would strictly (all-things-considered) have a stronger interest in being in the other state. So, if there’s an arrow which points out of the current outcome A, then A is dominated by the outcome B the arrow points to, from the “point of view” of outcome A (or the point of view of the individuals in outcome A). If an outcome is (all-things-considered) dominated by another in this way and there's no stronger (all-things-considered) domination in the exact opposite direction, it is not stable (in something like a game theoretic sense), and it’s impermissible to choose over the more strongly dominating outcome. You should not choose an outcome A that, after occurring, you’d expect (the aggregate of individuals in the outcome) to have a stronger interest in another outcome B, if the individuals in that outcome B don't have an even stronger interest in A. This is basically regret minimization, and it’s pretty similar to antifrustrationism.
Other outcomes may of course be impermissible to choose for other reasons.
Here are the arguments for each arrow, pertaining only to the interests of the individual whose existence we're considering:
If the individual would come to exist and have a negative existence, it would have been better for them to not have come to exist at all in the first place (e.g. they may prefer to not to have come to exist). Because they would exist, they have interests. So, there is an arrow from Negative existence (bottom left) to Nonexistence. This defends claim 1 in the Procreation Asymmetry.
Nonexistent (and never existing) individuals have no interests in coming to exist, since they have no interests at all, because they have no point of view (and, if they don't come to exist, they have no preference to exist). So, there is no arrow from Nonexistence (the right side of the graph) to Existence (the left side of the graph). This defends claim 2 in the Procreation Asymmetry.
Existing individuals also have an interest to be at least as well off as they are now or strictly better, so there are arrows going up in Existence.
Furthermore, those who do exist often prefer to continue to exist and may have interests in continuing to exist, so ending an individual's existence may be a harm. However, once they stop existing, they no longer prefer to exist and any such interest disappears.
Our obligations to others are, in this way, conditional on their existence: if they don’t come to exist, they have no stake in the outcome. Existence can be understood in a narrow or wide view way: in a narrow view, identities matter, but in a wide view, they don’t. The Nonidentity Problem is a classical illustration of the differences between these two views.
This argument is essentially comparative. Whether or not you must give up the independence of irrelevant alternatives (which is the claim that the ranking of outcomes does not depend on what outcomes are possible, so if A < B given for a given set of outcomes, then A < B for any set of outcomes) depends on which other outcomes you’re allowed to compare to, e.g. only realistically probable alternatives or also hypothetical ones. If you accept the asymmetry and are willing to give up the independence of irrelevant alternatives, you may find the following utilitarian approach for comparing outcomes intuitive:
His approach is to measure the harm in an outcome with respect to a set of outcomes as the sum of the differences between an individual’s maximum utility and their utility in the given outcome, using 0 if they don’t exist in the given outcome, and then minimize this harm. He then extends this approach to cases with different individuals (different identities) through “saturating counterpart relations”, taking a wide view of existence.
Meacham’s approach also solves the Nonidentity Problem, and avoids the Repugnant Conclusion and a specific principle of Antinatalism, which I define as the claim that it would be better for anyone who would come to exist with non-maximal utility to not come to exist at all, other things being equal. Note, however, that other things are rarely equal in practice, since in realistic scenarios individuals will have effects on others. This particular Antinatalism principle does not imply that we generally should not have children in practice (even considering the probabilistic risks of them having a bad life), because there may be other reasons why having a child is good, e.g. they may have positive impacts on the wellbeing of others.
Otherwise, if you aren’t willing to give up the independence of irrelevant alternatives, but still accept the asymmetry, then this leads to the principle of Antinatalism defined above. This is because for any individual with non-maximal utility, we can imagine the same individual (or, if the Nonidentity Problem is solved, another hypothetical individual) with higher utility, and that would be strictly better (a weak assumption), and by 2 of the Procreation Asymmetry, no better than not coming to exist at all, so by transitivity (and the completeness of the order, both weak assumptions), not coming to exist at all would be strictly better.
This post is co-authored with Ben Garfinkel. It is cross-posted from the CEA blog. A PDF version can be found here.
Summary: Some strategic decisions available to the effective altruism m...
TL;DR: I'm releasing a website that ranks philanthropists according to EA principles and research, and allows users to re-rank the list using their own assumptions. I'd like feedback and help making it better. I'd especially like ideas for how to make the results more trustworthy. Funding may be available.
Crossposted to LessWrong.
...
TL;DR: Marginal Victories is a new initiative to provide 1:1 career advising, opportunities, and resources for people exploring high-leverage U.S. democracy preservation and political work. Built by impact-oriented people doing pro-democracy work, the early MVP is now up at marginalvictories.org. Fill out the 10-minute form now to receive these resources as they become available over the next few...
It looks like "Conditional Reasons and the Procreation Asymmetry" was published in 2020, so it's even more recent than the other articles you mention (you probably read a pre-publication draft).
I just read the paper and I really liked it.