This is not a bad idea—in high school we had to run a charity fundraiser and I pushed us to support Pratham, which at the time was the only GiveWell-recommended charity in India. But I think supporting animal charities outright makes more sense, since you can probably have a bigger impact that way. The main concern with animal charities is the evidence base is not as strong, although it seems that the case for corporate outreach is pretty straightforward (which is probably why Open Phil made a grant in this area).
There could also be some overlap in helping developing countries reduce their birth rates, e.g. contracepitve availability, young female education, etc. This could increase per capita gdp, decrease child mortality and improve other metrics, without increasing total meat consumption exponentially. Perhaps this could be more of a focus area in EA.
I agree that family planning could become a new EA focus area. There is a facebook group to discuss family planning charities from a EA perspective. Giving What We Can has a great blog post on research around adolescent pregnancy. Development Media International is conducting a RCT to test whether radio programs can create demand for family planning.
Relevant: Repugnant Interventions, a great talk by Hilary Graves about how it’s plausibly inconsistent to support both life-saving interventions and family planning.
From the outset, I’m not sure how prevalent and unified a response there could be, because different EAs have different starting beliefs. Some of them agree that meat consumption is a serious issue and some do not, and this will necessarily change the way they look at the problem. I guess you could refine it and say that we need an “EAs-who-donate-to-poverty-alleviation” response, and maybe this is what you really mean, but if you constrain it like that then they don’t necessarily need a response because, by prioritizing donations to poverty over donations to animal causes, they’ve already demonstrated that they don’t care about animals as much, and if they did care about animals enough to be worried about this then they really ought to have been donating to animal welfare charities in the first place. For this reason, everything that I’m about to write regarding the relationship between wealth and meat consumption is 90% a mental exercise (and 10% a concern about how much to cooperate and support other people’s altruism). But more on that later.
The argument presented in Compassion by the Pound leads me to think, at the very least, that certain systems of animal agriculture might offer net-positive lives for the animals within them, albeit nowhere near what they deserve, and that outreach strategies that reduce demand for meat might not necessarily be net-positive. If we assume that the author’s welfare numbers are correct for a brief moment, it is entirely possible that reducing demand for meat would actually prevent some animals from being alive when they would’ve benefited from more enjoyment than suffering overall.
On average, it seems that animals’ lives on farms are negative. Chickens seem to exist in the greatest numbers, so their welfare dominates. Sure, there is a probability that animal lives are better than this author believes. But there is an equal probability that animal lives are even worse than this author believes. So we should just stick with the point estimate, which is negative.
Regarding shaky ethical assumptions: it seems a lot more plausible, across multiple ethical theories, that it’s wrong to create animals with dubious well-being in order to kill them. That it would be good to farm animals with uninteresting but net-pleasant lives is only something that a subset of utilitarians (and perhaps a few others) would endorse. I believe it’s plausible and quite likely right, but it does seem to require more controversial ethical assumptions.
For example, health interventions that save lives, like bed nets, may lower human fertility in the long run, resulting in lower meat consumption.
Well, they’ll reduce population growth, which will slow the rate of increase in meat consumption. But this just means that naive estimates of future increases in meat consumption are going to be slight underestimates rather than significant underestimates.
In this case, for saving lives in the developing world to be ethically bad, you would need to prove that over the next several generations, 1) people in countries with GiveWell-supported charities will be eating some high percentage of factory farmed meat, 2) factory farmed meat will produce net negative utility for animals, and that the moral weight of the harm to animals is greater than the humans saved in the first place.
I don’t think these are very contentious, or at least no more contentious than some of the countervailing ideas you are presenting.
I think we should put reasonable probability (ex. at least 10-20%) that we’ll develop widespread meat alternatives and/or have a high number of successful farm animal welfare campaigns that influence meat consumption in the developing world.
Well, the fact that EAs might find a solution isn’t necessarily what should change our evaluation, right? The probability that anyone or anything will find a solution to meat consumption in the developing world is what should move us, because that’s what really drives a reduction in expected increases in meat consumption.
So I’d guess that 10% is an astonishingly optimistic figure for thinking that EAs will even substantially reduce the meat consumption of an entire continent, but the probability that some kind of meat replacement will become very popular in Africa for any reason might come somewhere close to that. Note, however, that meat consumption is very important for improving the health of developing communities in Africa, and they will have a significant social demand for it (there are one or two sources on this which I can dig up if you want). Still, any big change in this issue will take a long time to spread, so even if something like this does happen, there is all the marginal meat consumption in the near and medium term future.
Honestly, this is what looks to me as a very contentious causal chain. That doesn’t make it wrong, but to be fair we’ll just have to add back comparable contentious causal chains which also point in favor of worrying about the meat eater problem. I think we should also consider the possibility that increased meat production in Africa will facilitate a large new industry with significant economies of scale, enabling a decrease in prices which will outprice meat alternatives and further entrench meat in African culture.
Fellow Effective Altruists might also argue that we should simply give to wherever has the highest expected utility, rather than stretch our donations across causes or organizations. In theory, I agree with this approach. However, I struggle to see how we can measure the impact of our donations against one another to the degree required to make this judgment with any degree of confidence, especially in light of uncertainties like the meat eater problem and far future concerns. For example, would you be able to say with greater than 80-90% confidence that a donation to the Against Malaria Foundation has greater expected value than to the Humane League or Future of Humanity Institute?
This is sort of the whole idea behind effective altruism anyway… Yes, it’s difficult, but if you think that animals are important enough that an additional couple weeks of suffering on farms matters in comparison to a $1,000 increase in annual income, then you should be donating to animal charities in the first place. If you compare the $3,000 it takes to save a life from malaria to the amount of advocacy and reform you can push for $3,000, it turns out that the animal-welfare opportunity cost of saving people is much greater than the animal-welfare direct cost of saving people. This comes from my own highly pessimistic, watered-down mental adjustments of the old vegan outreach ad calculations, but you can do it with other methods of animal charity and probably get similar results. So I think there is very little room in rational decision space for this concern to change where someone donates to—maybe if you care just a little bit about animals in comparison to humans, and are really unsure about the tradeoff, then the meat eater problem will tip the scales in favor of animal welfare donations.
The good thing about bringing up this issue, which is the reason I would like to thank you for this post, is that it seems to help people sort out their beliefs and come closer to the realization that if they care about this then they should have been donating to animal welfare all along.
What I don’t want, of course, is for EAs to feel inclined to care less about the meat eater problem and animal charities simply because they’ve already been donating to poverty alleviation. This would be an understandable and predictable form of cognitive bias. So, check yourselves for bias, everyone!
One clarification: Norwood’s view, as indicated in the table above, is that broiler chickens (raised for meat) have a welfare score of +3 which means they have lives worth living. Norwood does believe that the breeders (parents) of broilers have a welfare score of −4 (better off dead), but the ratio of breeders to broilers is 1 to 144 so his conclusion is that eating chicken increases animal welfare.
That differs from egg laying hens. Norwood gives caged hens, which currently represent the vast majority of the egg laying hens in the US, a welfare score of −8.
FWIW, I tend to disagree with Norwood’s views about broiler chickens and believe they are probably better off dead.
Okay, interesting. I was thinking more about Brian Tomasik’s numbers on quantifying suffering. Yeah, if it is the case that chickens tend to enjoy lives then the sign of meat consumption in Africa could very well flip, as the majority of the marginal animal-days of farming are taken by chickens.
I agree we won’t ever have a single response, but that’s not my intention. I just think this is an important enough problem that far more EAs should be taking it seriously and considering it as they donate/work (the EA response).
On chickens—yes, most chicken lives currently are probably very net-negative. However, the authors’ numbers say cage-free and market (non-breeder) chickens raised for meat live net-positive lives. If you disagree with the authors’ numbers, that’s a totally fair argument and I’d love to hear it. However, given the huge movement towards cage-free just in the past year and the numbers above, we may have many chickens living net-positive lives in the immediate future. This seems important to me as we discuss predictions about the next 50-100 years.
Re: shaky ethical assumptions: I agree that this is controversial and a view not held by many people. I’d love to hear arguments about why this ethical view is not correct!
Thanks for your link, I meant to put it in my post but forgot.
I was using a broader “we” as in “humanity will develop meat alternatives,” not that any particular Effective Altruist will do it. I don’t much care who does it.
However it’s worth keeping in mind that this is intensive agriculture in Africa. I’m not personally informed on what their factory farms are like, and we really don’t know how it could turn out in the long run. They might not be happy to adopt the same regulations that there are in the developed world (or they could be better, I suppose).
Unfortunately, I don’t have any relevant papers off the top of my head regarding ethics, but the repugnant conclusion, natalism and antinatalism, and animal rights would be good general areas to read into.
If you’re going to make an argument like this you should specify how you weight different animals, including wild ones, and include welfare estimates for wild animals. I have some discussion in this post.
For example, health interventions that save lives, like bed nets, may lower human fertility in the long run, resulting in lower meat consumption.
This is not obvious. See David Roodman’s write-up. Using the criteria expressed elsewhere in the post (total QALY-counting without a specified weighting scheme, direct effects and indirect effects on meat consumption but ignoring longer-term flow-through effects), this would be an argument for health interventions being bad, along Repugnant Conclusion lines.
A more practically sound approach, to me, seems to be to consider where the distribution of donations in the Effective Altruist movement generally falls and allocate funding as “bets” according to how much an organization could use money at the margins and its expected value. Under this line of thinking, I believe we may be under-prioritizing animal organizations, far future research, and meta-organizations like Giving What We Can.
This could work smoothly in an expected value framework for interventions with strong diminishing returns over the range of money EA can move, and would be recommended by relevant moral uncertainty or moral pluralism (this is relevant for many Open Philanthropy activities), although the cutoffs for each intervention will depend on relative weightings/credences/shares of influence.
However, it doesn’t work in the same way for interventions with near-constant returns over larger ranges. For instance, scaled-up cash transfers could absorb hundreds of billions of dollars per year with fairly steady marginal returns.
Yep, I think the discussion around how much we value different animal lives is pretty central to this. I think it deserves a post on its own—perhaps that’s the next thing I’ll write!
I think you’re right in theory about interventions with constant returns, but I’m not sure many interventions actually behave this way. To take GiveDirectly, I see one of the most large (potential) benefits being that developing countries may begin cash transfer systems after seeing GiveDirectly’s success. To that end, $50 million looks very different from $150 to $300 in how quickly countries will hear about their successes, how much media attention GiveDirectly receives, etc. It’s probably very impossible to predict where these cut-offs are—I’m just trying to highlight that optimizing our donations is of course what we should aim for, but pretty hard when many of the benefits come from policy changes from a diverse set of actors.
I don’t know how much the following argument works, but it’s possible that making people care about animals will increase their concern about the welfare of the world’s poorest people. The specific study that makes me think this could be true is The Brain Functional Networks Associated to Human and Animal Suffering Differ among Omnivores, Vegetarians and Vegans (source).
Thanks for the post. I’m somewhat less confident in the meat-eater problem being a problem as a result of it, maybe for different reasons though. I still think that it is overall a problem, however. I’ll just put my initial thoughts below.
It’s also plausible that interventions that raise incomes, like deworming, have a lower impact on meat consumption because they don’t raise the overall number of humans that would be eating meat over their entire lifetime.
The effect of raising income itself will still tend to increase meat consumption, though. There was another helpful post recently on the forum which attempted to quantify the effect of economic growth on meat consumption. Although, it’s plausible that interventions that raise incomes and contribute to increased education, such as deworming, could not only not raise the number of humans eating meat, but could also reduce the number of humans who would have otherwise existed, if education and particularly female education does lead to lower fertility, though I don’t think that this lowering of fertility would outweigh the increased amount of meat being eaten.
Also, while life-saving interventions may have no effect on or even lower fertility in the long-run, there’s also some evidence that interventions against malaria, for instance, may raise incomes too, which would lead to more meat being eaten. But, then again, more education as a result of lack of disruption due to malaria prevention could lead to lower fertility in the long run.
I’m less sure of other systems of animal agriculture where welfare standards are higher
Though, factory-farming is the dominant method of animal agriculture in the UK too, and likely in Europe as a whole. I’m also not convinced that animal welfare standards in developing countries will be significantly better even today, and I think that the hypothesis that factory-farming is only going to grow as incomes and populations grow is strong.
Even if I’m wrong about the meat eater problem, we can improve the chances we’ll solve it with investments in animal organizations today
I agree with this, and I also agree with the conclusion that EA should be directing more resources towards animal advocacy, because it does appear to be quite human-centred despite the commitment to impartiality. The possibility that lab-grown meat could ensure that the meat-eater problem is not as big of a problem in the future is also an interesting one, and hopefully one which will be realised.
Again, I would agree that this makes the meat-eater problem somewhat less of a concern, but it also means that potential short-term increases in fertility, which are plausible as a result of global health interventions as the report you cite states, are more important than long-run decreases in fertility—decreasing fertility in the long run is less likely to matter as more people in the long-run will impact less, in expectation, on meat-eating due to the ever increasing probability of lab-grown meat becoming widely or near-fully adopted.
I also liked the idea of “working more in India” as a compromise solution.
However, I’d still disagree that we should split our donations—I would endorse the view that we should maximise expected utility and favour your option 1 of donating solely to animal charities (or future animal suffering), and I wouldn’t say that this relies on implausible causal chains either. While I have downshifted my confidence in the meat-eater problem being a thing, I still think that it’s more likely to be a thing than not. And, I would say that the amount of suffering inflicted upon non-human animals as a result of meat-eating is greater than the amount of human suffering we could alleviate. So, if we’re sufficiently worried about the meat-eater problem, chances are that our donations would align best with our values if we donated solely to animal charities, and vice-versa.
Thanks, Vidur! I didn’t make it totally clear, but I don’t think that individuals should split their donations. The main argument that I’m trying to make is that the distribution of our donations across the EA movement are heavily human-centered, and that’s a mistake based on expected value. I didn’t want to dive too deep into this but that’s the claim I was trying to make.
Broadly speaking, I’d like to see a much higher proportion of our dollars go to animal organizations. I could see this being fixed from a decent sized group of people moving their donations over or a major organization like OPP fixing it.
One quick response: The people whose lives are saved by the Against Malaria Foundation are usually too poor to afford much meat (Malawians consume 25x less meat than Americans), and farmed animals in developing countries plausibly have better lives than those in developed countries at the moment, so I’m not very concerned about an immediate negative impact of AMF.
On the other hand, an increased population of Malawi now could lead to increased meat consumption in the future if the nation becomes wealthier in the future.
The effects of increased population on wild animal suffering are also important, which leads me to be unsure about whether increasing populations is net good. I can’t immediately find a good link for this, but this is an alright starting point.
This essay by Brian Tomasik addresses this question further, looking at the overall impact of human activities on wild-animal suffering, and includes the effect of factory-farming in the analysis too. Whilst human impact on the environment may lead to a net reduction in wild-animal suffering (if you think that the lives of wild-animals are significantly net-negative), the people whose lives are saved by the Against Malaria Foundation also have little impact on the environment, so also have little impact on the reduction of wild-animal suffering.
It’s plausible that those saved from malaria have lower-than-average environmental impact, but their impact is not trivial. This section mentions some ways in which poverty might actually increase a person’s environmental impact.
This section discusses AMF as a potential way to reduce insect suffering. I added a paragraph specifically about Malawi because Buck mentioned that country. I’m interested in finding someone to research the net impact of AMF on insect suffering more thoroughly. :)
Thanks, Buck! You’re definitely right in the short term. However, I think that the time frame we need to operate on is likely at least 1 generation, considering that’s about the amount of time someone will be eating meat if their life is saved by a health intervention. Perhaps even longer depending on the extent to which health interventions influence fertility.
You’re absolutely right on wild animal suffering, and I hope someone with more knowledge on this will chime in. I didn’t include it in the article because I find most of the arguments extremely speculative, but it’s not fair to entirely remove that from the equation, either.
This problem generalizes to other areas as well: increasing human populations almost certainly mean increased environmental strain, for example. I don’t want to digress into that too much, but I will say that I think you brought up a good response to this sort of critique: we shouldn’t assume too much about future causal chains. There is a great deal of awareness at the moment of both factory farming and climate change, and it is entirely possible that the effects of one additional human will be far lower than current states would project them to be.
Off the top of my head, I believe there are about 8-10 billion factory farmed animals consumed every year in the US—roughly 35 per person. Most of these are chickens, and I think we can say they lead net-negative lives (I simply do not believe non-breeder chickens are at a 4). There is no doubt that increasing the size of the US population would substantially increase the number of factory farmed animals, and an increase of 35 factory farmed animals per human is far from trivial. However, the good news is that the vast majority of lives extended by EAs are not in the US, and thus the increase in meat consumption their longer lives will produce is likely going to be sourced from animals with lives that are probably worth living.
I absolutely agree regarding animal charities. The potential is quite large simply due to the number of animals involved in factory farming and the quite poor conditions many of them live in. There are probably a relatively small number of humans who are actually living net negative lives. There are undoubtedly billions of animals who are living such lives. This alone should probably be a very important consideration for EAs.
Thanks, Zach! Point 3 seems especially important to me and something that I may highlight more in future articles.
I wanted to dive more into point 1 about environmental stress, but it’s so difficult! I genuinely don’t know whether eating beef is unethical or not—Compassion by the Pound’s welfare numbers are pretty positive for cows but their climate impact is much larger than chickens/pigs. I think it’s fairly clear that veal is bad, but hamburgers? It’s one question I’d like to see someone dive into.
I think that humans are the only chance to all other animal species to survive their normal way to extinction. Most species exist around 4 million years and all life on Earth will die off in next 1 billion year or earlier because of Sun’s rising luminosity. But humans start to resurrect extinct species and will safe animal life if humanity will be able to colonise the Galaxy.
That is why doing good to humans and to preventing x-risks is the best way we could help animals.
Who said we will preserve wild nature in its present form? We will re-engineer it to eliminate animal suffering while enhancing positive animal experience and wild nature’s aesthetic appeal.
The number of people who want to re-engineer nature is currently much, much smaller than the number of dedicated conservationists. It is a fringe view that basically only effective altruists support, and not even all EAs. I see no reason to believe that humans will ever modify wild animals to be more happy. Humans might eventually destroy most habitats, however.
This perspective is widely restated but I’m not sure this is supportable by argument:
Isn’t it almost certain that humans will eventually destroy most existing habitats? We’ve already destroyed in the vicinity of half, right, by proportion of land?
Most social change is a fringe interest initially. If we have good reasons to care about animal welfare in the abstract, then the interest in this may continue to increase. If one does not think have confidence in these arguments, then instead, mightn’t one want to take moral uncertainty or moral pluralism more seriously?
Creating animal environments in new planets in which they would not naturally live will involve a significantly different discussion compared to the treatment of wild animals.
In general, are you trying to generalise from humans’ treatment of animals in the 21st century to humans’ treatment of animals, in which modification of animals is very difficult, to an environment for the next thousands of years, that may or may not be in a simulated environment, grown in vitro, genetically engineered, et cetera, in which modification may be less difficult. If this is the generalisation that you are trying to make, then more thorough argumentation is needed. Obviously I’m also trying to generalise to the future, but the current naturalistic biases aren’t an obviously particularly relevant factor when the future situation is fleshed out more concretely.
Although most people currently don’t want to alter nature, in any circumstances that we worry about, people will have different capacities, that shape different views, about something that would no longer be aptly called “nature”, and so we need to reason differently about what to expect.
Most people will help wild animal if they see it in trouble now.
Anyway I think that we could create nanoimplants, which will be able to prevent suffering of wild animals by blocking excessive pain in case of death or injury. But these implants will not change the ways of their ordinary life, so natural life will look like almost the same.
I am also would vote for resurrection of all sentient life, starting from humans, but also including animals from most complex one to less complex. Probably future AI could do it.
Most people will help wild animal if they see it in trouble now.
Most people I know don’t really care about wild animals. The ones who do tend to be environmentalists who care more about preserving nature as it is than maximizing welfare. Do you have any evidence that most people actually would support reducing wild animal suffering?
I am also would vote for resurrection of all sentient life, starting from humans, but also including animals from most complex one to less complex.
I don’t see what the purpose of that would be. If you’re a classical total utilitarian, you should just determine what the happiest species is and make more of that (or make utilitronium). If you’re not a classic total utilitarian, then why would creating new beings help?
In an intergalactic civilisation, why would one expect un-augmented wild animals to represent any significant fraction of all life? There are strong incentives to use resources for human flourishing. Not a rhetorical question but I can’t think of any reason more compelling than the econ incentives. Notwithstanding language used in past pieces by pessimists, they don’t tend to contest this.
Most people don’t generally just want to reduce animal suffering without regard to preferences and happiness of animals. Nor is it reasonable to want purely that, when people have good arguments for other moral perspectives.
It’s also generally agrees that there’s not much reason to expect that there to be a lot of suffering animals in the long run, compared to the amount of animal flourishing, if we’re able to travel to new planets, make synthetic meats, create awesome entertainment and scientific experiments without their use, et cetera.
If terraforming other planets involves spreading nature, then it would be very bad according to any nonspeciesist utilitarian framework. So bad that it makes increasing x-risk look like a good idea.
I’m not convinced that we are morally obligated to create beings who will experience happiness/preference satisfaction. That just seems absurd, because nonexistence doesn’t deprive anyone of anything. On the other hand, creating beings who experience suffering definitely seems bad.
More intense lives will be able to be engineered, on expectation, for a longer time period, at a higher density, and across a larger space, on expectation, through biological augmentation or virtual reality, than through nature. So
trrraforming is a red herring here, because most (approximately all) human and animal experience will be engineered by biotech in the long run.
most (approximately all) human and animal experience will be engineered by biotech in the long run
You’re making a very strong claim about something that will happen in the future that has never happened in the past based on speculation about what’s technologically feasible and on what the people with power will want to do. Maybe you’re right but you seem really overconfident here.
I mean “on expectation” as in it’s at least slightly more likely than not, based on what little.wr currently know, but I’m still very interested in new evidence.
Do you think it is likely that humans will run sentient simulations in the future? It could be that wild animal brain simulations and “suffering subroutines” dominate future expected utility calculations.
Sure, any subroutines could exist in the future. In artificial worlds, the range of possible experience should be much larger than presently. The incentives for researching and developing entertainment should by much larger than for engineering psychological harm. Generalisations from the natural world wouldn’t necessarily follow to simulations, but on the inside view, net flourishing is expected.
It looks like you’re subscribing to a person-affecting philosophy, whereby you say potential future humans aren’t worthy of moral consideration because they’re not being deprived, but bringing them into existence would be bad because they would (could) suffer.
I think this is arbitrarily asymmetrical, and not really compatible with a total utilitarian framework. I would suggest reading the relevant chapter in Nick Beckstead’s thesis ‘On the overwhelming importance of shaping the far future’, where I think he does a pretty good job at showing just this.
Another tack: You don’t have to include the creation of new beings in the calculation. There are plenty who already exist. How many orders of magnitude do you expect between the intensity/eneegy-ensity/numver of natural experiences compared to expected synthetic ones. There seem to be strong arguments that make the natural experiences irrelevant.
But the majority of beings that already exist are wild animals with negative lives. I’m not sure what you’re trying to argue here. Do you mean something like “already will exist”?
Even if there’s only a small chance of people achieving life extension, the humans currently alive have a pretty long expected life, big enough to make wild animals currently alive less relevant (though concievably there could be other cohorts even more important).
Not that attempts to make a principled distinction between moral treatment of ‘deprivation’ and ‘fulfilment’ or ‘exist’ and ‘already will exist’ don’t seem to be particularly philosophically satisfying or seem to go much beyond restatement of a personal intuition, anyway, in so far as I can tell.
Do you seriously believe that there is a non-negligible chance that any human alive today will be alive in, let’s say, 2000 years? That sounds like wishful thinking to me.
I don’t think classical total utilitarianism is the correct theory of population ethics. If you do, I suppose breeding and wireheading a bunch of rats is a great way to help the world, but that just seems silly.
Lol, I’m glad I was a salient example of someone with silly beliefs =P. Just doing my part to push the Overton window.
Strictly speaking, I expect there could be beings that are a lot happier than rats (or any other current living thing), so we should really breed those instead.
No, the negative preference utilitarian version wants to prevent Bob from surviving to have children and numerous future generations (even if their lives are all great overall), so strongly as to overwhelm any horrible thing that happens to Bob, including painfully killing him (or torturing him for thousands of years, if that was instrumentally useful) to prevent his procreation.
This is why we need to implement my own theory, “Negative-Leaning Average Preference Prioritarianism with Time-Discounted Utility for Future Generations and with Extra Points Awarded for Minimizing the Variance of Utilities Among the Present Generation.”
In this particular example, Rawlsianism will also point to making the least well-off better before creating new lives or killing existing ones (though it’s quite possible your stomach wouldn’t be as badly off as the kids with schistosomiasis).
Wishful thinking? Hardly. A 1% chance of being alive in 2000 years is too unlikely psychologically useful to me, but a smallet chance of being around forever is mathematically decisive for the specific stated purpose of making utility calculations.
Alternative approaches to population are even worse because instead of the so-called “repugnant conclusion” (it can be debated) you get the “sadistic conclusion”. Alternatives are not only less principled, but worse (1). None can ever be satisfactory (2).
If you dispense with utilitarian approaches (or at least allow alternatives to contribute), then terraforming or the lack of it is less of a focus.
A Pascal’s mugging is an intentional move by an actor where they are probably deceiving you. The fact that something has a low probability of huge payoff doesn’t make it a mugging and doesn’t imply that we should ignore it.
I don’t think moral uncertainty is a real problem. The slope isn’t uneven; I just believe that suffering is worse than most EAs do, but it would still be a straight line. I also do not support creating new happy beings instead of helping those who already exist and are suffering.
I don’t think “moral uncertainty” is something that can be solved, or even a legitimate meta-ethical problem. You can’t compare how bad something is across multiple ethical theories. Is 1 violation of rights = 1 utilon? There’s also the possibility that the correct ethical theory hasn’t even been discovered yet, and we don’t have any idea what it would say.
Cool. Interestingly, twice you’ve surprised me by endorsing a position that I thought you were repudiating. A straight line in terms of experience and value is exactly what I think of by symmetric utilitarianism, just as puzzling over this question is just what I imagine by thinking moral uncertainty is a problem. The idea that the correct ethical theory hasn’t been discovered yet, if there is such a thing, seems to be the most important source of uncertainty of all to me, though it is rarely discussed.
I believe in a symmetry for people who already exist, but I also think empirically that many common sources of suffering are far worse than the common sources of happiness are good. For people who don’t exist, I don’t see how creating more happy people is good. The absence of happiness is not bad. This is where I think there is an asymmetry.
I don’t even understand what it would mean for an ethical theory to be correct. Does that mean it is hardwired into the physical constants of the universe? I guess I’m sort of a non-cognitivist.
Right, but is that for sources of happiness and suffering that are common among all people who will exist across all time? Because almost all of the people who will exist (irrespective of your actions) don’t currently.
There’s a difficulty that I guess you’d be sensitive to, in that it’s hard to distinguish the absence of happiness from the presence of suffering and vice versa. The difference between the two is not hardwired into the physical constants of the universe, if that is a phrasing that you might be sympathetic to, though no snark is intended.
If you’re non-cognitivist, then you could ask whether you “should” (even rationally or egoistically) act according to your moral perspective. If you choose to live out your values by some description, for some reason, then they’re not going to be purely represented by any ethical theory anyway, and it’d be unclear to me why you’d want to simplify your intuitions in that way.
If you don’t have a child, you are not decreasing your nonexistent offspring’s welfare/preference satisfaction. Beings who do not exist do not have preferences and cannot suffer. Once they exist (and become sentient), their preferences and welfare matter. This may not be hardcoded into the universe, but it’s not hard to distinguish between having a child and not having one.
I meant within one person. If you believe that there is a fundamental difference between intrapersonal and interpersonal comparisons, then you’re going to run into a wall trying to define persons… It doesn’t seem to me that this really checks out, putting aside the question of why one would want simple answers here as a non-cognitivist.
It’s also generally agrees that there’s not much reason to expect that there to be a lot of suffering animals in the long run, compared to the amount of animal flourishing, if we’re able to travel to new planets, make synthetic meats, create awesome entertainment and scientific experiments without their use, et cetera.
This is not a consensus. Most people don’t care about suffering in nature or equivalent ecosystems. There is also no consensus that we should outlaw animal use even if we invent fully functional substitutes.
I personally think it’s naive to expect more flourishing than suffering even in humans. Just because a culture is technologically advanced doesn’t mean they won’t torture the defenseless on a large scale. I expect this to happen to humans and posthumans frequently. There is nothing in the universe that will prevent this.
It’s not a consensus but none of the authors/researchers who I would.expect to argue this actually do expect the expected amount of animal suffering to outweigh the amount of animal flourishing in the long run. On the other hand, dozens of prominent researchers including Shulman, Beckstead, Wiblin and Bostrom, many of whom are hard-nosed utilitarians, have come to conclude the opposite.
What I’m looking for in a credible assessment of this question is for people to think about what kinds of worlds we might see. Then the trick is to focus not on the worlds with a particular salient scenario in them, but the ones that are most durable, with large scope and large populstion. Such worlds will be outliers in the sense that they are not natural anymore, we might live much longer, there may not be a meaningful category of a “human” anymore. There may no-longer be multiple living entities anymore but perhaps just one. Or there may be much better ways to understand the experiences of other beings. We may have a very different approach to morality. It may be possible to create accurate models of a being without simulating their emotions. We may have a better understanding of the mechanics of emotions. Et cetera et cetera.
That kind of thinking, sharpened with an empirical approach that takes note of past improvements in technology and welfare, is needed to thoroughly investigate this issue, not a “single issue” presumption about a topic to one’s personal interests, however interesting that topic may seem.
The fact that their analyses include a wide range of topics, rather than focusing on confirming and emphasising specific hypotheses is encouraging, and the fact that a large number of credible people have arrived at similar conclusions from widely varying perspectives is the best possible sign.
Typo(s?) 7th full paragraph, not counting the bolded sentences
“However, it’s likely that people in the developing world will eat not eat as much meat as people in the developing world for quite some time and that some fraction of the meat they do eat will not come from factory farms.”
Really interesting article, thanks for writing it! I’m especially intrigued by the concept of the possible + net utility of certain forms of factory farming, as this is an idea I have not encountered before, and have updated based on it.
It’s also plausible that interventions that raise incomes, like deworming, have a lower impact on meat consumption because they don’t raise the overall number of humans that would be eating meat over their entire lifetime.
I’m a bit concerned with this argument. There’s research showing that people in developing countries eat more meat as their income increases. So if the goal is to optimize for lower meat consumption, it is beneficial to keep people’s incomes lower.
Now, I am not saying that we should keep people’s incomes lower. In fact, I’m a strong supporter of GiveDirectly and other organizations that increase people’s incomes in developing countries. However, we cannot blind ourselves to the fact that the consequence of increasing people’s incomes is an increase in meat consumption, and we have to count that as a variable in the cost-benefit analysis.
Let’s work more in India
While in a way that makes sense, we have to remember that in India, beef and pork meat are rarely eaten for religious reasons (Hindus don’t eat beef, Muslims don’t eat pork). So chicken meat is going to be eaten more there. This is a concern that we need to include in our calculations.
EA could probably stand to give a much higher proportion of its money to animal charities
As someone concerned with movement-building from the perspective of how the EA movement looks to outsiders, we should consider the costs and benefits from a PR perspective of this move. I’m not saying it’s a bad or good move, just raising it as an issue to consider.
(Hindus don’t eat beef, Muslims don’t eat pork). So chicken meat is going to be eaten more there.
This is plausible in principle, but it seems that in practice, India has quite low per-capita poultry consumption, although this is partly because India is so poor even relative to China.
It turns out that India is the biggest exporter of beef and the biggest producer of milk. This page shows India as having the world’s biggest cattle population. Of course, some of the causal responsibility for this belongs to beef consumption in the importing countries.
possible + net utility of certain forms of factory farming
As someone concerned with movement-building from the perspective of how the EA movement looks to outsiders, we should consider the costs and benefits from a PR perspective of this move.
I think the proportions in which EAs donate have very little to do with public perception. For example, lots of people like to complain about how EAs care too much about AI safety even though only about 1% of EAs’ money goes toward AI safety.
Good point! I will update my point to being concerned with how the EA movement talks about its priorities in donations, versus how much people actually donate.
Thanks, Gleb! I definitely struggle with the PR aspect of this—it’s certainly a weird topic but one that I think matters a lot.
Definitely think that we should include increased meat consumption in our cost-effectiveness analysis for interventions that increase income. My guess is that this amount is much smaller than for interventions that save lives, like bed nets, but that’s certainly an open question.
I agree with Brian’s remarks on chicken consumption in India—it didn’t seem the case when I looked at the data.
This is not a bad idea—in high school we had to run a charity fundraiser and I pushed us to support Pratham, which at the time was the only GiveWell-recommended charity in India. But I think supporting animal charities outright makes more sense, since you can probably have a bigger impact that way. The main concern with animal charities is the evidence base is not as strong, although it seems that the case for corporate outreach is pretty straightforward (which is probably why Open Phil made a grant in this area).
There could also be some overlap in helping developing countries reduce their birth rates, e.g. contracepitve availability, young female education, etc. This could increase per capita gdp, decrease child mortality and improve other metrics, without increasing total meat consumption exponentially. Perhaps this could be more of a focus area in EA.
I agree that family planning could become a new EA focus area. There is a facebook group to discuss family planning charities from a EA perspective. Giving What We Can has a great blog post on research around adolescent pregnancy. Development Media International is conducting a RCT to test whether radio programs can create demand for family planning.
Relevant: Repugnant Interventions, a great talk by Hilary Graves about how it’s plausibly inconsistent to support both life-saving interventions and family planning.
Both AMF and family planning improve lives, so in that sense they are compatible.
To the people who downvote us here (or perhaps just one guy with 2 accounts):
Feel free to provide actual constructive criticism. Or solve the problem of global poverty, child mortality, and animal suffering on your own.
From the outset, I’m not sure how prevalent and unified a response there could be, because different EAs have different starting beliefs. Some of them agree that meat consumption is a serious issue and some do not, and this will necessarily change the way they look at the problem. I guess you could refine it and say that we need an “EAs-who-donate-to-poverty-alleviation” response, and maybe this is what you really mean, but if you constrain it like that then they don’t necessarily need a response because, by prioritizing donations to poverty over donations to animal causes, they’ve already demonstrated that they don’t care about animals as much, and if they did care about animals enough to be worried about this then they really ought to have been donating to animal welfare charities in the first place. For this reason, everything that I’m about to write regarding the relationship between wealth and meat consumption is 90% a mental exercise (and 10% a concern about how much to cooperate and support other people’s altruism). But more on that later.
On average, it seems that animals’ lives on farms are negative. Chickens seem to exist in the greatest numbers, so their welfare dominates. Sure, there is a probability that animal lives are better than this author believes. But there is an equal probability that animal lives are even worse than this author believes. So we should just stick with the point estimate, which is negative.
Regarding shaky ethical assumptions: it seems a lot more plausible, across multiple ethical theories, that it’s wrong to create animals with dubious well-being in order to kill them. That it would be good to farm animals with uninteresting but net-pleasant lives is only something that a subset of utilitarians (and perhaps a few others) would endorse. I believe it’s plausible and quite likely right, but it does seem to require more controversial ethical assumptions.
Well, they’ll reduce population growth, which will slow the rate of increase in meat consumption. But this just means that naive estimates of future increases in meat consumption are going to be slight underestimates rather than significant underestimates.
I don’t think these are very contentious, or at least no more contentious than some of the countervailing ideas you are presenting.
In my quick analysis of the issue (http://effective-altruism.com/ea/rl/quantifying_the_impact_of_economic_growth_on_meat/) I had a source saying that much (probably most) of new animal agriculture in the developing world is intensive (http://www.worldwatch.org/node/1826). I estimated 17 days of animal suffering, including 14 days of poultry suffering, for an increase of $1,000 in average per-capita GDP.
Well, the fact that EAs might find a solution isn’t necessarily what should change our evaluation, right? The probability that anyone or anything will find a solution to meat consumption in the developing world is what should move us, because that’s what really drives a reduction in expected increases in meat consumption.
So I’d guess that 10% is an astonishingly optimistic figure for thinking that EAs will even substantially reduce the meat consumption of an entire continent, but the probability that some kind of meat replacement will become very popular in Africa for any reason might come somewhere close to that. Note, however, that meat consumption is very important for improving the health of developing communities in Africa, and they will have a significant social demand for it (there are one or two sources on this which I can dig up if you want). Still, any big change in this issue will take a long time to spread, so even if something like this does happen, there is all the marginal meat consumption in the near and medium term future.
Honestly, this is what looks to me as a very contentious causal chain. That doesn’t make it wrong, but to be fair we’ll just have to add back comparable contentious causal chains which also point in favor of worrying about the meat eater problem. I think we should also consider the possibility that increased meat production in Africa will facilitate a large new industry with significant economies of scale, enabling a decrease in prices which will outprice meat alternatives and further entrench meat in African culture.
This is sort of the whole idea behind effective altruism anyway… Yes, it’s difficult, but if you think that animals are important enough that an additional couple weeks of suffering on farms matters in comparison to a $1,000 increase in annual income, then you should be donating to animal charities in the first place. If you compare the $3,000 it takes to save a life from malaria to the amount of advocacy and reform you can push for $3,000, it turns out that the animal-welfare opportunity cost of saving people is much greater than the animal-welfare direct cost of saving people. This comes from my own highly pessimistic, watered-down mental adjustments of the old vegan outreach ad calculations, but you can do it with other methods of animal charity and probably get similar results. So I think there is very little room in rational decision space for this concern to change where someone donates to—maybe if you care just a little bit about animals in comparison to humans, and are really unsure about the tradeoff, then the meat eater problem will tip the scales in favor of animal welfare donations.
Carl Shulman also objected to this argument along similar lines: http://effective-altruism.com/ea/rl/quantifying_the_impact_of_economic_growth_on_meat/60t
The good thing about bringing up this issue, which is the reason I would like to thank you for this post, is that it seems to help people sort out their beliefs and come closer to the realization that if they care about this then they should have been donating to animal welfare all along.
What I don’t want, of course, is for EAs to feel inclined to care less about the meat eater problem and animal charities simply because they’ve already been donating to poverty alleviation. This would be an understandable and predictable form of cognitive bias. So, check yourselves for bias, everyone!
One clarification: Norwood’s view, as indicated in the table above, is that broiler chickens (raised for meat) have a welfare score of +3 which means they have lives worth living. Norwood does believe that the breeders (parents) of broilers have a welfare score of −4 (better off dead), but the ratio of breeders to broilers is 1 to 144 so his conclusion is that eating chicken increases animal welfare.
That differs from egg laying hens. Norwood gives caged hens, which currently represent the vast majority of the egg laying hens in the US, a welfare score of −8.
FWIW, I tend to disagree with Norwood’s views about broiler chickens and believe they are probably better off dead.
Okay, interesting. I was thinking more about Brian Tomasik’s numbers on quantifying suffering. Yeah, if it is the case that chickens tend to enjoy lives then the sign of meat consumption in Africa could very well flip, as the majority of the marginal animal-days of farming are taken by chickens.
Thanks, Kbog! Responding to a few claims:
I agree we won’t ever have a single response, but that’s not my intention. I just think this is an important enough problem that far more EAs should be taking it seriously and considering it as they donate/work (the EA response).
On chickens—yes, most chicken lives currently are probably very net-negative. However, the authors’ numbers say cage-free and market (non-breeder) chickens raised for meat live net-positive lives. If you disagree with the authors’ numbers, that’s a totally fair argument and I’d love to hear it. However, given the huge movement towards cage-free just in the past year and the numbers above, we may have many chickens living net-positive lives in the immediate future. This seems important to me as we discuss predictions about the next 50-100 years.
Re: shaky ethical assumptions: I agree that this is controversial and a view not held by many people. I’d love to hear arguments about why this ethical view is not correct!
Thanks for your link, I meant to put it in my post but forgot.
I was using a broader “we” as in “humanity will develop meat alternatives,” not that any particular Effective Altruist will do it. I don’t much care who does it.
Okay, I understand.
However it’s worth keeping in mind that this is intensive agriculture in Africa. I’m not personally informed on what their factory farms are like, and we really don’t know how it could turn out in the long run. They might not be happy to adopt the same regulations that there are in the developed world (or they could be better, I suppose).
Unfortunately, I don’t have any relevant papers off the top of my head regarding ethics, but the repugnant conclusion, natalism and antinatalism, and animal rights would be good general areas to read into.
If you’re going to make an argument like this you should specify how you weight different animals, including wild ones, and include welfare estimates for wild animals. I have some discussion in this post.
This is not obvious. See David Roodman’s write-up. Using the criteria expressed elsewhere in the post (total QALY-counting without a specified weighting scheme, direct effects and indirect effects on meat consumption but ignoring longer-term flow-through effects), this would be an argument for health interventions being bad, along Repugnant Conclusion lines.
This could work smoothly in an expected value framework for interventions with strong diminishing returns over the range of money EA can move, and would be recommended by relevant moral uncertainty or moral pluralism (this is relevant for many Open Philanthropy activities), although the cutoffs for each intervention will depend on relative weightings/credences/shares of influence.
However, it doesn’t work in the same way for interventions with near-constant returns over larger ranges. For instance, scaled-up cash transfers could absorb hundreds of billions of dollars per year with fairly steady marginal returns.
Yep, I think the discussion around how much we value different animal lives is pretty central to this. I think it deserves a post on its own—perhaps that’s the next thing I’ll write!
I think you’re right in theory about interventions with constant returns, but I’m not sure many interventions actually behave this way. To take GiveDirectly, I see one of the most large (potential) benefits being that developing countries may begin cash transfer systems after seeing GiveDirectly’s success. To that end, $50 million looks very different from $150 to $300 in how quickly countries will hear about their successes, how much media attention GiveDirectly receives, etc. It’s probably very impossible to predict where these cut-offs are—I’m just trying to highlight that optimizing our donations is of course what we should aim for, but pretty hard when many of the benefits come from policy changes from a diverse set of actors.
Good points, Carl!
I don’t know how much the following argument works, but it’s possible that making people care about animals will increase their concern about the welfare of the world’s poorest people. The specific study that makes me think this could be true is The Brain Functional Networks Associated to Human and Animal Suffering Differ among Omnivores, Vegetarians and Vegans (source).
Thanks for the post. I’m somewhat less confident in the meat-eater problem being a problem as a result of it, maybe for different reasons though. I still think that it is overall a problem, however. I’ll just put my initial thoughts below.
The effect of raising income itself will still tend to increase meat consumption, though. There was another helpful post recently on the forum which attempted to quantify the effect of economic growth on meat consumption. Although, it’s plausible that interventions that raise incomes and contribute to increased education, such as deworming, could not only not raise the number of humans eating meat, but could also reduce the number of humans who would have otherwise existed, if education and particularly female education does lead to lower fertility, though I don’t think that this lowering of fertility would outweigh the increased amount of meat being eaten.
Also, while life-saving interventions may have no effect on or even lower fertility in the long-run, there’s also some evidence that interventions against malaria, for instance, may raise incomes too, which would lead to more meat being eaten. But, then again, more education as a result of lack of disruption due to malaria prevention could lead to lower fertility in the long run.
Though, factory-farming is the dominant method of animal agriculture in the UK too, and likely in Europe as a whole. I’m also not convinced that animal welfare standards in developing countries will be significantly better even today, and I think that the hypothesis that factory-farming is only going to grow as incomes and populations grow is strong.
I agree with this, and I also agree with the conclusion that EA should be directing more resources towards animal advocacy, because it does appear to be quite human-centred despite the commitment to impartiality. The possibility that lab-grown meat could ensure that the meat-eater problem is not as big of a problem in the future is also an interesting one, and hopefully one which will be realised.
Again, I would agree that this makes the meat-eater problem somewhat less of a concern, but it also means that potential short-term increases in fertility, which are plausible as a result of global health interventions as the report you cite states, are more important than long-run decreases in fertility—decreasing fertility in the long run is less likely to matter as more people in the long-run will impact less, in expectation, on meat-eating due to the ever increasing probability of lab-grown meat becoming widely or near-fully adopted.
I also liked the idea of “working more in India” as a compromise solution.
However, I’d still disagree that we should split our donations—I would endorse the view that we should maximise expected utility and favour your option 1 of donating solely to animal charities (or future animal suffering), and I wouldn’t say that this relies on implausible causal chains either. While I have downshifted my confidence in the meat-eater problem being a thing, I still think that it’s more likely to be a thing than not. And, I would say that the amount of suffering inflicted upon non-human animals as a result of meat-eating is greater than the amount of human suffering we could alleviate. So, if we’re sufficiently worried about the meat-eater problem, chances are that our donations would align best with our values if we donated solely to animal charities, and vice-versa.
Thanks, Vidur! I didn’t make it totally clear, but I don’t think that individuals should split their donations. The main argument that I’m trying to make is that the distribution of our donations across the EA movement are heavily human-centered, and that’s a mistake based on expected value. I didn’t want to dive too deep into this but that’s the claim I was trying to make.
Broadly speaking, I’d like to see a much higher proportion of our dollars go to animal organizations. I could see this being fixed from a decent sized group of people moving their donations over or a major organization like OPP fixing it.
One quick response: The people whose lives are saved by the Against Malaria Foundation are usually too poor to afford much meat (Malawians consume 25x less meat than Americans), and farmed animals in developing countries plausibly have better lives than those in developed countries at the moment, so I’m not very concerned about an immediate negative impact of AMF.
On the other hand, an increased population of Malawi now could lead to increased meat consumption in the future if the nation becomes wealthier in the future.
The effects of increased population on wild animal suffering are also important, which leads me to be unsure about whether increasing populations is net good. I can’t immediately find a good link for this, but this is an alright starting point.
This essay by Brian Tomasik addresses this question further, looking at the overall impact of human activities on wild-animal suffering, and includes the effect of factory-farming in the analysis too. Whilst human impact on the environment may lead to a net reduction in wild-animal suffering (if you think that the lives of wild-animals are significantly net-negative), the people whose lives are saved by the Against Malaria Foundation also have little impact on the environment, so also have little impact on the reduction of wild-animal suffering.
Thanks for the link!
It’s plausible that those saved from malaria have lower-than-average environmental impact, but their impact is not trivial. This section mentions some ways in which poverty might actually increase a person’s environmental impact.
This section discusses AMF as a potential way to reduce insect suffering. I added a paragraph specifically about Malawi because Buck mentioned that country. I’m interested in finding someone to research the net impact of AMF on insect suffering more thoroughly. :)
Thanks, Buck! You’re definitely right in the short term. However, I think that the time frame we need to operate on is likely at least 1 generation, considering that’s about the amount of time someone will be eating meat if their life is saved by a health intervention. Perhaps even longer depending on the extent to which health interventions influence fertility.
You’re absolutely right on wild animal suffering, and I hope someone with more knowledge on this will chime in. I didn’t include it in the article because I find most of the arguments extremely speculative, but it’s not fair to entirely remove that from the equation, either.
This problem generalizes to other areas as well: increasing human populations almost certainly mean increased environmental strain, for example. I don’t want to digress into that too much, but I will say that I think you brought up a good response to this sort of critique: we shouldn’t assume too much about future causal chains. There is a great deal of awareness at the moment of both factory farming and climate change, and it is entirely possible that the effects of one additional human will be far lower than current states would project them to be.
Off the top of my head, I believe there are about 8-10 billion factory farmed animals consumed every year in the US—roughly 35 per person. Most of these are chickens, and I think we can say they lead net-negative lives (I simply do not believe non-breeder chickens are at a 4). There is no doubt that increasing the size of the US population would substantially increase the number of factory farmed animals, and an increase of 35 factory farmed animals per human is far from trivial. However, the good news is that the vast majority of lives extended by EAs are not in the US, and thus the increase in meat consumption their longer lives will produce is likely going to be sourced from animals with lives that are probably worth living.
I absolutely agree regarding animal charities. The potential is quite large simply due to the number of animals involved in factory farming and the quite poor conditions many of them live in. There are probably a relatively small number of humans who are actually living net negative lives. There are undoubtedly billions of animals who are living such lives. This alone should probably be a very important consideration for EAs.
Overall, great article!
Thanks, Zach! Point 3 seems especially important to me and something that I may highlight more in future articles.
I wanted to dive more into point 1 about environmental stress, but it’s so difficult! I genuinely don’t know whether eating beef is unethical or not—Compassion by the Pound’s welfare numbers are pretty positive for cows but their climate impact is much larger than chickens/pigs. I think it’s fairly clear that veal is bad, but hamburgers? It’s one question I’d like to see someone dive into.
I think that humans are the only chance to all other animal species to survive their normal way to extinction. Most species exist around 4 million years and all life on Earth will die off in next 1 billion year or earlier because of Sun’s rising luminosity. But humans start to resurrect extinct species and will safe animal life if humanity will be able to colonise the Galaxy. That is why doing good to humans and to preventing x-risks is the best way we could help animals.
We don’t want to rewild or spread wild animals to other realms.
http://reducing-suffering.org/will-space-colonization-multiply-wild-animal-suffering/
http://reducing-suffering.org/applied-welfare-biology-wild-animal-advocates-focus-spreading-nature/
Who said we will preserve wild nature in its present form? We will re-engineer it to eliminate animal suffering while enhancing positive animal experience and wild nature’s aesthetic appeal.
The number of people who want to re-engineer nature is currently much, much smaller than the number of dedicated conservationists. It is a fringe view that basically only effective altruists support, and not even all EAs. I see no reason to believe that humans will ever modify wild animals to be more happy. Humans might eventually destroy most habitats, however.
This perspective is widely restated but I’m not sure this is supportable by argument:
Isn’t it almost certain that humans will eventually destroy most existing habitats? We’ve already destroyed in the vicinity of half, right, by proportion of land?
Most social change is a fringe interest initially. If we have good reasons to care about animal welfare in the abstract, then the interest in this may continue to increase. If one does not think have confidence in these arguments, then instead, mightn’t one want to take moral uncertainty or moral pluralism more seriously?
Creating animal environments in new planets in which they would not naturally live will involve a significantly different discussion compared to the treatment of wild animals.
In general, are you trying to generalise from humans’ treatment of animals in the 21st century to humans’ treatment of animals, in which modification of animals is very difficult, to an environment for the next thousands of years, that may or may not be in a simulated environment, grown in vitro, genetically engineered, et cetera, in which modification may be less difficult. If this is the generalisation that you are trying to make, then more thorough argumentation is needed. Obviously I’m also trying to generalise to the future, but the current naturalistic biases aren’t an obviously particularly relevant factor when the future situation is fleshed out more concretely.
Although most people currently don’t want to alter nature, in any circumstances that we worry about, people will have different capacities, that shape different views, about something that would no longer be aptly called “nature”, and so we need to reason differently about what to expect.
Most people will help wild animal if they see it in trouble now. Anyway I think that we could create nanoimplants, which will be able to prevent suffering of wild animals by blocking excessive pain in case of death or injury. But these implants will not change the ways of their ordinary life, so natural life will look like almost the same. I am also would vote for resurrection of all sentient life, starting from humans, but also including animals from most complex one to less complex. Probably future AI could do it.
Most people I know don’t really care about wild animals. The ones who do tend to be environmentalists who care more about preserving nature as it is than maximizing welfare. Do you have any evidence that most people actually would support reducing wild animal suffering?
I don’t see what the purpose of that would be. If you’re a classical total utilitarian, you should just determine what the happiest species is and make more of that (or make utilitronium). If you’re not a classic total utilitarian, then why would creating new beings help?
In an intergalactic civilisation, why would one expect un-augmented wild animals to represent any significant fraction of all life? There are strong incentives to use resources for human flourishing. Not a rhetorical question but I can’t think of any reason more compelling than the econ incentives. Notwithstanding language used in past pieces by pessimists, they don’t tend to contest this.
There are some arguments for potential future suffering here:
http://foundational-research.org/risks-of-astronomical-future-suffering/
That would increase animal suffering. We want to decrease it.
Most people don’t generally just want to reduce animal suffering without regard to preferences and happiness of animals. Nor is it reasonable to want purely that, when people have good arguments for other moral perspectives.
It’s also generally agrees that there’s not much reason to expect that there to be a lot of suffering animals in the long run, compared to the amount of animal flourishing, if we’re able to travel to new planets, make synthetic meats, create awesome entertainment and scientific experiments without their use, et cetera.
If terraforming other planets involves spreading nature, then it would be very bad according to any nonspeciesist utilitarian framework. So bad that it makes increasing x-risk look like a good idea.
I’m not convinced that we are morally obligated to create beings who will experience happiness/preference satisfaction. That just seems absurd, because nonexistence doesn’t deprive anyone of anything. On the other hand, creating beings who experience suffering definitely seems bad.
More intense lives will be able to be engineered, on expectation, for a longer time period, at a higher density, and across a larger space, on expectation, through biological augmentation or virtual reality, than through nature. So trrraforming is a red herring here, because most (approximately all) human and animal experience will be engineered by biotech in the long run.
Arguments not downvotes, please!
You’re making a very strong claim about something that will happen in the future that has never happened in the past based on speculation about what’s technologically feasible and on what the people with power will want to do. Maybe you’re right but you seem really overconfident here.
I mean “on expectation” as in it’s at least slightly more likely than not, based on what little.wr currently know, but I’m still very interested in new evidence.
Do you think it is likely that humans will run sentient simulations in the future? It could be that wild animal brain simulations and “suffering subroutines” dominate future expected utility calculations.
Sure, any subroutines could exist in the future. In artificial worlds, the range of possible experience should be much larger than presently. The incentives for researching and developing entertainment should by much larger than for engineering psychological harm. Generalisations from the natural world wouldn’t necessarily follow to simulations, but on the inside view, net flourishing is expected.
It looks like you’re subscribing to a person-affecting philosophy, whereby you say potential future humans aren’t worthy of moral consideration because they’re not being deprived, but bringing them into existence would be bad because they would (could) suffer.
I think this is arbitrarily asymmetrical, and not really compatible with a total utilitarian framework. I would suggest reading the relevant chapter in Nick Beckstead’s thesis ‘On the overwhelming importance of shaping the far future’, where I think he does a pretty good job at showing just this.
Another tack: You don’t have to include the creation of new beings in the calculation. There are plenty who already exist. How many orders of magnitude do you expect between the intensity/eneegy-ensity/numver of natural experiences compared to expected synthetic ones. There seem to be strong arguments that make the natural experiences irrelevant.
But the majority of beings that already exist are wild animals with negative lives. I’m not sure what you’re trying to argue here. Do you mean something like “already will exist”?
Even if there’s only a small chance of people achieving life extension, the humans currently alive have a pretty long expected life, big enough to make wild animals currently alive less relevant (though concievably there could be other cohorts even more important).
Not that attempts to make a principled distinction between moral treatment of ‘deprivation’ and ‘fulfilment’ or ‘exist’ and ‘already will exist’ don’t seem to be particularly philosophically satisfying or seem to go much beyond restatement of a personal intuition, anyway, in so far as I can tell.
Do you seriously believe that there is a non-negligible chance that any human alive today will be alive in, let’s say, 2000 years? That sounds like wishful thinking to me.
I don’t think classical total utilitarianism is the correct theory of population ethics. If you do, I suppose breeding and wireheading a bunch of rats is a great way to help the world, but that just seems silly.
Lol, I’m glad I was a salient example of someone with silly beliefs =P. Just doing my part to push the Overton window.
Strictly speaking, I expect there could be beings that are a lot happier than rats (or any other current living thing), so we should really breed those instead.
When you’re advocating a reductio ad absurdum, I do wonder if that pushes the overton window backwards.
Bob: “Ouch, my stomach hurts.”
Classical total utilitarian: “Don’t worry! Wait while I create more happy people to make up for it.”
Average utilitarian: “Never fear! Let me create more people with only mild stomach aches to improve the average.”
Egalitarian: “I’m sorry to hear that. Here, let me give everyone else awful stomach aches too.”
...
Negative utilitarian: “Here, take this medicine to make your stomach feel better.”
The medicine is a lethal dose of sedatives.
Negative preference utilitarianism avoids that problem.
No, the negative preference utilitarian version wants to prevent Bob from surviving to have children and numerous future generations (even if their lives are all great overall), so strongly as to overwhelm any horrible thing that happens to Bob, including painfully killing him (or torturing him for thousands of years, if that was instrumentally useful) to prevent his procreation.
This is why we need to implement my own theory, “Negative-Leaning Average Preference Prioritarianism with Time-Discounted Utility for Future Generations and with Extra Points Awarded for Minimizing the Variance of Utilities Among the Present Generation.”
We’ll call it NLAPPTDUFGEPAMVUAPG.
As a classical total utilitarian, I think we should both give medicine to make your stomach feel better AND create more happy people.
Or ideally, we should create people who don’t get stomach aches in the first place.
To the extent this example has force, it seems to push towards prioritarianism rather than negative utilitarianism.
In this particular example, Rawlsianism will also point to making the least well-off better before creating new lives or killing existing ones (though it’s quite possible your stomach wouldn’t be as badly off as the kids with schistosomiasis).
Wishful thinking? Hardly. A 1% chance of being alive in 2000 years is too unlikely psychologically useful to me, but a smallet chance of being around forever is mathematically decisive for the specific stated purpose of making utility calculations.
Alternative approaches to population are even worse because instead of the so-called “repugnant conclusion” (it can be debated) you get the “sadistic conclusion”. Alternatives are not only less principled, but worse (1). None can ever be satisfactory (2).
If you dispense with utilitarian approaches (or at least allow alternatives to contribute), then terraforming or the lack of it is less of a focus.
1) interactive guide to population ethics, ben west http://people.su.se/~guarr/Texter/The%20Impossibility%20of%20a%20Satisfactory%20Population%20Ethics%20in%20Descriptive%20and%20Normative%20Approaches%20to%20Human%20Behavior%202011.pdf 2).http://people.su.se/~guarr/Texter/The%20Impossibility%20of%20a%20Satisfactory%20Population%20Ethics%20in%20Descriptive%20and%20Normative%20Approaches%20to%20Human%20Behavior%202011.pdf
This seems like a textbook case of a Pascal’s mugging.
I would describe my ethical view as negative-leaning (or perhaps asymmetric)), but still broadly utilitarian.
A Pascal’s mugging is an intentional move by an actor where they are probably deceiving you. The fact that something has a low probability of huge payoff doesn’t make it a mugging and doesn’t imply that we should ignore it.
How do you involve moral uncertainty or moral pluralism?
How do you set the scale for happiness and suffering on which moral value is supposed to slope unevenly? (1)
http://www.amirrorclear.net/academic/ideas/negative-utilitarianism/images/graph.png
I don’t think moral uncertainty is a real problem. The slope isn’t uneven; I just believe that suffering is worse than most EAs do, but it would still be a straight line. I also do not support creating new happy beings instead of helping those who already exist and are suffering.
Are you completely certain that you should act according to your moral perspective?
I don’t think “moral uncertainty” is something that can be solved, or even a legitimate meta-ethical problem. You can’t compare how bad something is across multiple ethical theories. Is 1 violation of rights = 1 utilon? There’s also the possibility that the correct ethical theory hasn’t even been discovered yet, and we don’t have any idea what it would say.
Cool. Interestingly, twice you’ve surprised me by endorsing a position that I thought you were repudiating. A straight line in terms of experience and value is exactly what I think of by symmetric utilitarianism, just as puzzling over this question is just what I imagine by thinking moral uncertainty is a problem. The idea that the correct ethical theory hasn’t been discovered yet, if there is such a thing, seems to be the most important source of uncertainty of all to me, though it is rarely discussed.
I believe in a symmetry for people who already exist, but I also think empirically that many common sources of suffering are far worse than the common sources of happiness are good. For people who don’t exist, I don’t see how creating more happy people is good. The absence of happiness is not bad. This is where I think there is an asymmetry.
I don’t even understand what it would mean for an ethical theory to be correct. Does that mean it is hardwired into the physical constants of the universe? I guess I’m sort of a non-cognitivist.
Right, but is that for sources of happiness and suffering that are common among all people who will exist across all time? Because almost all of the people who will exist (irrespective of your actions) don’t currently.
There’s a difficulty that I guess you’d be sensitive to, in that it’s hard to distinguish the absence of happiness from the presence of suffering and vice versa. The difference between the two is not hardwired into the physical constants of the universe, if that is a phrasing that you might be sympathetic to, though no snark is intended.
If you’re non-cognitivist, then you could ask whether you “should” (even rationally or egoistically) act according to your moral perspective. If you choose to live out your values by some description, for some reason, then they’re not going to be purely represented by any ethical theory anyway, and it’d be unclear to me why you’d want to simplify your intuitions in that way.
If you don’t have a child, you are not decreasing your nonexistent offspring’s welfare/preference satisfaction. Beings who do not exist do not have preferences and cannot suffer. Once they exist (and become sentient), their preferences and welfare matter. This may not be hardcoded into the universe, but it’s not hard to distinguish between having a child and not having one.
I meant within one person. If you believe that there is a fundamental difference between intrapersonal and interpersonal comparisons, then you’re going to run into a wall trying to define persons… It doesn’t seem to me that this really checks out, putting aside the question of why one would want simple answers here as a non-cognitivist.
This is not a consensus. Most people don’t care about suffering in nature or equivalent ecosystems. There is also no consensus that we should outlaw animal use even if we invent fully functional substitutes.
I personally think it’s naive to expect more flourishing than suffering even in humans. Just because a culture is technologically advanced doesn’t mean they won’t torture the defenseless on a large scale. I expect this to happen to humans and posthumans frequently. There is nothing in the universe that will prevent this.
It’s not a consensus but none of the authors/researchers who I would.expect to argue this actually do expect the expected amount of animal suffering to outweigh the amount of animal flourishing in the long run. On the other hand, dozens of prominent researchers including Shulman, Beckstead, Wiblin and Bostrom, many of whom are hard-nosed utilitarians, have come to conclude the opposite.
What I’m looking for in a credible assessment of this question is for people to think about what kinds of worlds we might see. Then the trick is to focus not on the worlds with a particular salient scenario in them, but the ones that are most durable, with large scope and large populstion. Such worlds will be outliers in the sense that they are not natural anymore, we might live much longer, there may not be a meaningful category of a “human” anymore. There may no-longer be multiple living entities anymore but perhaps just one. Or there may be much better ways to understand the experiences of other beings. We may have a very different approach to morality. It may be possible to create accurate models of a being without simulating their emotions. We may have a better understanding of the mechanics of emotions. Et cetera et cetera.
That kind of thinking, sharpened with an empirical approach that takes note of past improvements in technology and welfare, is needed to thoroughly investigate this issue, not a “single issue” presumption about a topic to one’s personal interests, however interesting that topic may seem.
Nick Bostrom et al. could be affected by confirmation bias and optimism bias.
The fact that their analyses include a wide range of topics, rather than focusing on confirming and emphasising specific hypotheses is encouraging, and the fact that a large number of credible people have arrived at similar conclusions from widely varying perspectives is the best possible sign.
Typo(s?) 7th full paragraph, not counting the bolded sentences
“However, it’s likely that people in the developing world will eat not eat as much meat as people in the developing world for quite some time and that some fraction of the meat they do eat will not come from factory farms.”
Really interesting article, thanks for writing it! I’m especially intrigued by the concept of the possible + net utility of certain forms of factory farming, as this is an idea I have not encountered before, and have updated based on it.
I’m a bit concerned with this argument. There’s research showing that people in developing countries eat more meat as their income increases. So if the goal is to optimize for lower meat consumption, it is beneficial to keep people’s incomes lower.
Now, I am not saying that we should keep people’s incomes lower. In fact, I’m a strong supporter of GiveDirectly and other organizations that increase people’s incomes in developing countries. However, we cannot blind ourselves to the fact that the consequence of increasing people’s incomes is an increase in meat consumption, and we have to count that as a variable in the cost-benefit analysis.
While in a way that makes sense, we have to remember that in India, beef and pork meat are rarely eaten for religious reasons (Hindus don’t eat beef, Muslims don’t eat pork). So chicken meat is going to be eaten more there. This is a concern that we need to include in our calculations.
As someone concerned with movement-building from the perspective of how the EA movement looks to outsiders, we should consider the costs and benefits from a PR perspective of this move. I’m not saying it’s a bad or good move, just raising it as an issue to consider.
Overall, very good article, upvoted!
This is plausible in principle, but it seems that in practice, India has quite low per-capita poultry consumption, although this is partly because India is so poor even relative to China.
It turns out that India is the biggest exporter of beef and the biggest producer of milk. This page shows India as having the world’s biggest cattle population. Of course, some of the causal responsibility for this belongs to beef consumption in the importing countries.
Another possible benefit of animal (or at least cattle) farming is lowering wild-insect populations.
Updated on India and beef exporting, thanks for the numbers!
I think the proportions in which EAs donate have very little to do with public perception. For example, lots of people like to complain about how EAs care too much about AI safety even though only about 1% of EAs’ money goes toward AI safety.
Good point! I will update my point to being concerned with how the EA movement talks about its priorities in donations, versus how much people actually donate.
Thanks, Gleb! I definitely struggle with the PR aspect of this—it’s certainly a weird topic but one that I think matters a lot.
Definitely think that we should include increased meat consumption in our cost-effectiveness analysis for interventions that increase income. My guess is that this amount is much smaller than for interventions that save lives, like bed nets, but that’s certainly an open question.
I agree with Brian’s remarks on chicken consumption in India—it didn’t seem the case when I looked at the data.
The PR aspect is pretty nuanced. I think we’d need to do some market research to actually know how it will cash out.
Regarding meat consumption, this shouldn’t be hard to figure out in a detailed analysis.
Yup, I updated based on Brian’s points regarding chicken consumption in India.