(Edited to add: I’m speaking only in a personal capacity here, and not on behalf of my employer.)
If you try to account broadly for indirect effects and externalities, it’s not clear you’ve arrived at the right stance on abortion and population reduction. There are many effects to consider:
Withdrawing support for (or actively opposing) abortion and the reasoning here to do so could have impacts on norms/values around exploitation and treating others like mere means (or allowing them to be treated like mere means), like Richard suggests, so (EDIT) to make explicit, have other negative effects for women and girls (or others) in the near term. Norms more permissive of exploitation and treating others like mere means may make us more prone to conflict and less cooperative in general, and so more prone to catastrophic conflicts and missing opportunities for positive sum trades. Values in general via value lock-in and such effects on conflict and cooperation in particular could even have far-future effects.
Abortion and population reduction (for humans) could go either way for nonhuman animals. I’d guess they’re good for farmed animals by reducing animal farming and because I think farmed animals mostly have bad lives (even on symmetric ethical views, although my views are suffering-focused), and the sign for wild animals (whose populations would plausibly increase) will depend on your expectations about their average welfare. I’d guess this would be limited to impacts on Earth, but humans might colonize space themselves and bring nonhuman animals with them, and more humans colonizing space could mean more nonhuman animals brought with them.
Abortion and population reduction could mitigate climate change, which may have near-term effects on humans and nonhuman animals, as well as far-future effects.
Moral circle expansion towards fetuses could transfer to MCE for other (far future) minds with limited cognitive capacities or agency, or otherwise in situations similar to fetuses.
Economic and technological effects from increased population or reducing costs on children (in time or money), which could possibly even compound into the far future.
There are a few broad practical responses to all of this that I can think of:
Try harder to quantify and compare these effects (EDIT: in some cases, rough bounds can be useful enough), and try to ensure you’re capturing a relatively unbiased subset of indirect effects, and especially the largest ones. Use this information to
come to an overall stance on specific interventions and support them accordingly, and/or
Cluelessness, and look for interventions that are more robustly positive to support and promote instead. This would also mean withdrawing support for human life-saving interventions (except possibly those that reduce existential risk overall, possibly). Possibly promote the withdrawal of support for abortion-affecting and human population-affecting interventions, including family planning interventions, interventions that incentivize people to have more children, life-saving interventions, and abortion reduction interventions.
Also, another possible response is 3. ignoring indirect effects, but this seems pretty unprincipled/unjustified and prone to systematic error to me. Ignoring cross-worldview or cross-cause indirect effects may be okay as an approximation to the portfolio approach if it’s done within a portfolio of interventions across causes/worldviews, because it’s plausible negative indirect effects can be made up for through more targeted/leveraged interventions for those causes/worldviews in the portfolios.
However, when I think of the current total EA portfolio of interventions, I think it’s pretty plausible we aren’t making up for possibly negative wild animal effects because agricultural land use is huge and fishing has huge population effects, and (although I’m much less informed on the issue) I also worry about s-risks being increased.
Yes, that’s absolutely a relevant consideration. I think there are similar considerations regarding the effect of supporting abortion on the permissiveness of bad societal norms:
The dehumanization of those outside one’s moral circle (“it’s just a clump of cells/parasite”, “it’s just a beast”, “they won’t even exist for millions of years”)
The callous treatment of moral patients whose existence is inconvenient (“it’s my body; I’ll do whatever I want with that fetus”, “who cares? meat tastes good”)
The masking of disenfranchisement of unrecognized moral patients as “rights” of recognized moral patients (“reproductive rights”, “the right to eat whatever I want”)
Yes, animal welfare considerations likely weigh in favor of abortion. I’d go even further than your statement and say they plausibly dominate the welfare concerns of both the woman and the fetus. Of course, this consideration plausibly dominates the purpose of many human-centric interventions, and the implications are scary. (That doesn’t mean I won’t think them through, but there’s only so fast I can overturn my worldview!) I know you have strong opinions on this, and would love to get any recommendations on reads you think would be enlightening on the subject.
Yep! That should also weigh in favor of abortion, though I personally think other concerns substantially dominate it.
100%
Yes. I could be persuaded either way on the sign of this consideration, because abortion does increase economic output per capita, but having more people should increase gross economic output on the margin.
On your responses:
Very cool! I wasn’t aware of research on this, and it makes perfect sense, including avoiding “sector risk” (e.g. a portfolio of human-centered interventions could be totally dominated by farmed animal welfare considerations).
I’m still making up my mind on the implications of cluelessness, but I agree that it updates towards not taking much concrete action on abortion as an EA cause area.
I think the human impacts on wild animals are primarily through agricultural land use, fishing and climate change, and most of these are largely affected by human diets (although climate change possibly dominated by fossil fuel use). Maybe also environmental pollutants/contaminants/toxins and forestry (I haven’t really looked into these). I’d guess the effects from the land humans take up in cities, towns, villages, etc., is not significant compared to these, based on Our World in Data.
For moral weights across animals (including humans):
I don’t know if the above covers all the strongest arguments for humans mattering substantially more than nonhuman animals, and I’d guess it doesn’t cover many such arguments in much detail. I don’t know off the top of my head what to recommend.
(Edited to add: I’m speaking only in a personal capacity here, and not on behalf of my employer.)
If you try to account broadly for indirect effects and externalities, it’s not clear you’ve arrived at the right stance on abortion and population reduction. There are many effects to consider:
Withdrawing support for (or actively opposing) abortion and the reasoning here to do so could have impacts on norms/values around exploitation and treating others like mere means (or allowing them to be treated like mere means), like Richard suggests, so (EDIT) to make explicit, have other negative effects for women and girls (or others) in the near term. Norms more permissive of exploitation and treating others like mere means may make us more prone to conflict and less cooperative in general, and so more prone to catastrophic conflicts and missing opportunities for positive sum trades. Values in general via value lock-in and such effects on conflict and cooperation in particular could even have far-future effects.
Abortion and population reduction (for humans) could go either way for nonhuman animals. I’d guess they’re good for farmed animals by reducing animal farming and because I think farmed animals mostly have bad lives (even on symmetric ethical views, although my views are suffering-focused), and the sign for wild animals (whose populations would plausibly increase) will depend on your expectations about their average welfare. I’d guess this would be limited to impacts on Earth, but humans might colonize space themselves and bring nonhuman animals with them, and more humans colonizing space could mean more nonhuman animals brought with them.
Abortion and population reduction could mitigate climate change, which may have near-term effects on humans and nonhuman animals, as well as far-future effects.
Moral circle expansion towards fetuses could transfer to MCE for other (far future) minds with limited cognitive capacities or agency, or otherwise in situations similar to fetuses.
Economic and technological effects from increased population or reducing costs on children (in time or money), which could possibly even compound into the far future.
There are a few broad practical responses to all of this that I can think of:
Try harder to quantify and compare these effects (EDIT: in some cases, rough bounds can be useful enough), and try to ensure you’re capturing a relatively unbiased subset of indirect effects, and especially the largest ones. Use this information to
come to an overall stance on specific interventions and support them accordingly, and/or
hedge with a portfolio of interventions or do some other kind of worldview diversification/portfolio approach.
Cluelessness, and look for interventions that are more robustly positive to support and promote instead. This would also mean withdrawing support for human life-saving interventions (except possibly those that reduce existential risk overall, possibly). Possibly promote the withdrawal of support for abortion-affecting and human population-affecting interventions, including family planning interventions, interventions that incentivize people to have more children, life-saving interventions, and abortion reduction interventions.
Also, another possible response is 3. ignoring indirect effects, but this seems pretty unprincipled/unjustified and prone to systematic error to me. Ignoring cross-worldview or cross-cause indirect effects may be okay as an approximation to the portfolio approach if it’s done within a portfolio of interventions across causes/worldviews, because it’s plausible negative indirect effects can be made up for through more targeted/leveraged interventions for those causes/worldviews in the portfolios.
However, when I think of the current total EA portfolio of interventions, I think it’s pretty plausible we aren’t making up for possibly negative wild animal effects because agricultural land use is huge and fishing has huge population effects, and (although I’m much less informed on the issue) I also worry about s-risks being increased.
Hi Michael, great to hear from you!
Yes, that’s absolutely a relevant consideration. I think there are similar considerations regarding the effect of supporting abortion on the permissiveness of bad societal norms:
The dehumanization of those outside one’s moral circle (“it’s just a clump of cells/parasite”, “it’s just a beast”, “they won’t even exist for millions of years”)
The callous treatment of moral patients whose existence is inconvenient (“it’s my body; I’ll do whatever I want with that fetus”, “who cares? meat tastes good”)
The masking of disenfranchisement of unrecognized moral patients as “rights” of recognized moral patients (“reproductive rights”, “the right to eat whatever I want”)
Yes, animal welfare considerations likely weigh in favor of abortion. I’d go even further than your statement and say they plausibly dominate the welfare concerns of both the woman and the fetus. Of course, this consideration plausibly dominates the purpose of many human-centric interventions, and the implications are scary. (That doesn’t mean I won’t think them through, but there’s only so fast I can overturn my worldview!) I know you have strong opinions on this, and would love to get any recommendations on reads you think would be enlightening on the subject.
Yep! That should also weigh in favor of abortion, though I personally think other concerns substantially dominate it.
100%
Yes. I could be persuaded either way on the sign of this consideration, because abortion does increase economic output per capita, but having more people should increase gross economic output on the margin.
On your responses:
Very cool! I wasn’t aware of research on this, and it makes perfect sense, including avoiding “sector risk” (e.g. a portfolio of human-centered interventions could be totally dominated by farmed animal welfare considerations).
I’m still making up my mind on the implications of cluelessness, but I agree that it updates towards not taking much concrete action on abortion as an EA cause area.
On animal effects, I would recommend:
For farmed animal and wild animal effects and population sizes, with some emphasis on those related to human diets:
Many essays in https://reducing-suffering.org/#animals (although note that these are primarily from a suffering-focused and basically negative utilitarian perspective)
Maybe especially https://reducing-suffering.org/#humanitys_impact , https://reducing-suffering.org/vegetarianism-and-wild-animals/, https://reducing-suffering.org/trophic-cascades-caused-fishing/ (and others in https://reducing-suffering.org/#fishing ) and https://reducing-suffering.org/how-many-wild-animals-are-there/
http://reflectivedisequilibrium.blogspot.com/2013/07/vegan-advocacy-and-pessimism-about-wild.html
https://docs.google.com/document/d/1wMIa6bAn4rfCAzBAsKlHBH9X3gmo8pjn/edit from https://www.invinciblewellbeing.com/research
Various posts at https://forum.effectivealtruism.org/topics/meat-eater-problem
https://forum.effectivealtruism.org/posts/SvbZtETGenTkZni8C/where-does-most-of-the-suffering-from-eating-meat-come-from
I think the human impacts on wild animals are primarily through agricultural land use, fishing and climate change, and most of these are largely affected by human diets (although climate change possibly dominated by fossil fuel use). Maybe also environmental pollutants/contaminants/toxins and forestry (I haven’t really looked into these). I’d guess the effects from the land humans take up in cities, towns, villages, etc., is not significant compared to these, based on Our World in Data.
For moral weights across animals (including humans):
https://reducing-suffering.org/two-envelopes-problem-for-brain-size-and-moral-uncertainty/
https://forum.effectivealtruism.org/posts/848SgRAKpjbuBWkW7/why-might-one-value-animals-far-less-than-humans
RP’s moral weight sequence (disclaimer: I work at RP, but am not speaking for them here), some posts are still coming out.
https://reducing-suffering.org/is-brain-size-morally-relevant/
https://www.lesswrong.com/posts/2jTQTxYNwo6zb3Kyp/preliminary-thoughts-on-moral-weight
I don’t know if the above covers all the strongest arguments for humans mattering substantially more than nonhuman animals, and I’d guess it doesn’t cover many such arguments in much detail. I don’t know off the top of my head what to recommend.