Lots of agreement with the spirit of this post, but I'd like to gripe a bit about the actual numbers here:
* I think you're using Brian Tomasik's catfish estimate for the farmed fish multiplier rather than his salmon estimate or a weighted combination of the two, but I'd expect the salmon one to be closer to accurate - US catfish consumption is around half a pound per person per year,[1] while annual US salmon consumption per capita is around 2.5 pounds[2] of which around two thirds[3] is farmed (at least as of 2004), so with just those two datapoints I'd guess that typical farmed fish consumption looks around a third as bad as pure catfish numbers would imply.
* Weight seems like a bit of an unfair metric for evaluating how much someone's giving up - calories seems more proportional to how much of a person's diet things take up, which makes the amount of sacrifice here more like 50% than 20%.
* The "eliminate 95% of suffering" claim here only goes through if you eliminate eggs/poultry/farmed fish AND replace their former roles in your diet with exclusively vegan foods. Depending on what you substitute with, this could make the ethical savings substantially worse. In particular, if a reader scales up their beef, pork, and wild-caught fish consumption to match the same total meat intake, this will roughly double the suffering of their diet relative to having substituted with vegan options, which makes the savings more like 90% (or 90 -> 80%, if we use adjusted fish numbers as a baseline).
* More subjective, but I don't think the numbers Brian uses in the calculator are all that representative of typical moral weight consensus - e.g., I'm personally inclined to put pigs/cows at much larger than 1.1, and to downweight fish more significantly. The cow/chicken ratio of 2 is much less extreme than found in surveys like those mentioned in https://slatestarcodex.com/2019/05/01/update-to-partial-retraction-of-animal-value-and-neuron-number/, where ratios range from 6.5 to 13 depending on the survey. Most of these changes push in the direction of making the suffering difference between food sources appear less extreme.
I'm really surprised by the low amount of suffering due to pork: intensive pig farming (crates etc) is pretty horrific and pigs are among the most sentient and intelligent meat animals, to the point that they are (or used to be) studied for stress response to help model humans.
Perhaps it's because so little pork is relatively eaten by Americans? Would you advise that people living in pork-heavy countries (some Asian, Eastern European etc) should also exclude pork if they want to follow ameliatarian diet?
Also, what about lamb? I'm assuming the amount of suffering is comparable or lower than beef, though the animals are much smaller, so how would that compare? And presumably wild game meat and "farmed" venison is lower on suffering than beef?
[I'm in the UK, if it makes a difference]
PS. It's probably of zero interest to you, but since reading your previous post on meat related suffering, I've excluded farmed salmon (I ate no other farmed fish) completely from my diet and it wasn't anywhere near as hard as I thought.
You seem to be using "Certified Humane" as an umbrella term, but it is not; rather, it is (FWICT) a legitimate standard that actually does provide for a reasonable standard of animal welfare.
The link you provide to the Humane Society's website has this to say about the Certified Humane label:
"The Certified Humane standards include minimum space allowances, bedding material and environmental enrichment (e.g., hay bales, pecking blocks and perches for chickens) among dozens of other basic requirements for animal health and nutrition. Animals are never confined in cages or crates and are free to display natural behavior. Farms are inspected by trained auditors."
This does sound like "functionally meaningless" to me. I have also previously looked into the Certified Humane standard in more detail, and also thought it seemed legit. I would certainly like to know if that is not the case.
My understanding is that the US government standards all suck, but 3rd party standards (such as Certified Humane) may be legit. But (from my research) Canadian ones do not. I don't know about the UK, but I would like to.
That's interesting, the last time I read you on this topic you had written that while most egg labels are meaningless, that "pasture raised" was different and actually meaningful. Did the situation change or did your mind change?
Not sure if this is the case everywhere, but in many EU countries chickens' access to the outdoors is currently legally limited due to bird flu concerns.
My understanding is that virtually all eggs in the US are factory-farmed. In the UK (where I live), it is relatively easy to obtain "free-range" eggs (which is a regulated, protected term- it can't be applied by egg producers at will). Quite how strict the rules are, I'm not 100% sure, but the birds have to be allowed to walk around out of doors, and have access to water.
Chicken meat, however, is virtually all factored-farmed even in the UK. It is possible to find "free-range" chicken, but it's not easy and it costs 3-4 times the price. Any chicken-containing product or restaurant meal will almost certainly not use free-range chicken.
1. Farmed fish are often fed fishmeal, which is generally made from wild-caught fish. Since wild-caught fish are smaller and more numerous, plausibly the effects of eating farmed fish are dominated by effects on wild fish. How do you weigh this when making a specific recommendation about farmed fish?
2. If the reply to the above is that if the expected value on wild fish is unclear, since fish farming seems clearly bad, the expected value turns negative: It seems to me like a lot of demand for wild fish is “satisfied” by farmed fish (in that price increases in wild fish cause consumers to substitute to farmed fish if wild fish supply is relatively inelastic). So I wouldn’t be surprised if eating wild fish turns out to have very similar effects on increasing fish farming as eating farmed fish (perhaps with the exception of generally wild-caught species without obvious farmed substitutes, like anchovies). How different do you think the impacts on fish farming are from wild vs farmed fish consumption?
Hi Tejas, I really appreciate your analysis of the second-order effects. I think about these issues too, so I tried to get some data on the problems.
I got estimates for salmon feed from [1] and picked the two most numerous fish in a salmon's diet with this feed. My estimate was about 75 anchovies and 55 sprats [2] So the second-order effects are roughly 2 orders of magnitude greater. The source also mentions farmed insects in the feed, which would further increase the second-order effects.
The blog post only seemed to imply that wild caught fish are killed before they would die naturally, making wild-caught fishing neutral. I agree that your mentioned effects might push this into negative territory. I have a lot of uncertainty around wild fishing for other reasons, but it could still be better than the counterfactual of chicken and shrimp. Chicken has its own bad second-order effects, and shrimp could dominate on scale (250+ shrimp = 1 salmon in calories).
Wild-caught fish has hovered around 90 million tons since 1990 while aquaculture has grown roughly exponentially to 120 million tons [3]. This fits the inelastic supply and increased fish farming story well. I'm not sure where the chain of effects bottoms out: demand for wild-caught -> more aquaculture -> increased demand for fishmeal/oil -> more demand for smaller fish… But it seems quite bad given that the substitution away from wild fish is likely toward farmed fish, shrimp, and chicken.
[2] (1 anchovy / 0.002 kg anchovy oil)(25 kg anchovy oil / 100 kg fish oil)(0.1 kg fish oil / 1 kg feed)(1.2 kg feed / 1 kg salmon)(5 kg salmon) = 75 anchovies
(1 sprat / 0.001 kg sprat oil)(9.5 kg sprat oil / 100 kg fish oil)(0.1 kg fish oil / 1 kg feed)(1.2 kg feed / 1 kg salmon)(5 kg salmon) = 54 sprats
There are two other levers that could be pulled to influence animal suffering caused by our diets:
1. Eating less meat and other animal products. The modern US diet contains far more animal protein than is optimal for our health. On average, we could eat a smaller amount of beef and cheese and come out ahead. When I've done things like this (for weight loss reasons), I found that there was actually pleasure in anticipating special meals and treats. "Oh, hey, tomorrow is cheeseburger day! It's going to be delicious!" Eating meat less often also allows for spending more money on the meat we do eat.
2. Only eating certain meat products where you have an extremely high degree of confidence that they were raised humanely. As you mentioned in the post, backyard eggs is a classic example of this. But plenty of people in rural areas have access to local farms, or to neighbors who, say, work together to raise a pig. In this case, it's feasible to actually inspect the farm! And this approach doesn't require being a coastal US elite with disposable income. Raising chickens for eggs is a classic low-tech, low-capital food source.
At the upper limit, if you really liked eggs, I suppose it would be possible to organize a consumer cooperative to raise chickens under genuinely humane conditions. Consumer cooperatives are awesome. They're halfway between a non-profit and a regular business, and they're democratically managed by their members. (Once they reach a certain scale, this usually means electing a board.) One big difference between consumer cooperatives and regular business is that consumer cooperatives usually don't scale beyond the needs of a local region, because they have limited access to capital. But if someone lives in the big city, and would enjoy the occasional omelet, it might work.
None of these options will provide the quantity of meat provided by factory farming, and they'll be more expensive. Which brings me back to (1), eating less meat.
This is a really good idea Ozy, I have basically been following this diet for years without thinking that I needed a word for it. Going to start using it and see if it catches on!
Lots of agreement with the spirit of this post, but I'd like to gripe a bit about the actual numbers here:
* I think you're using Brian Tomasik's catfish estimate for the farmed fish multiplier rather than his salmon estimate or a weighted combination of the two, but I'd expect the salmon one to be closer to accurate - US catfish consumption is around half a pound per person per year,[1] while annual US salmon consumption per capita is around 2.5 pounds[2] of which around two thirds[3] is farmed (at least as of 2004), so with just those two datapoints I'd guess that typical farmed fish consumption looks around a third as bad as pure catfish numbers would imply.
* Weight seems like a bit of an unfair metric for evaluating how much someone's giving up - calories seems more proportional to how much of a person's diet things take up, which makes the amount of sacrifice here more like 50% than 20%.
* The "eliminate 95% of suffering" claim here only goes through if you eliminate eggs/poultry/farmed fish AND replace their former roles in your diet with exclusively vegan foods. Depending on what you substitute with, this could make the ethical savings substantially worse. In particular, if a reader scales up their beef, pork, and wild-caught fish consumption to match the same total meat intake, this will roughly double the suffering of their diet relative to having substituted with vegan options, which makes the savings more like 90% (or 90 -> 80%, if we use adjusted fish numbers as a baseline).
* More subjective, but I don't think the numbers Brian uses in the calculator are all that representative of typical moral weight consensus - e.g., I'm personally inclined to put pigs/cows at much larger than 1.1, and to downweight fish more significantly. The cow/chicken ratio of 2 is much less extreme than found in surveys like those mentioned in https://slatestarcodex.com/2019/05/01/update-to-partial-retraction-of-animal-value-and-neuron-number/, where ratios range from 6.5 to 13 depending on the survey. Most of these changes push in the direction of making the suffering difference between food sources appear less extreme.
[1]: https://www.aces.edu/blog/topics/farm-management/us-farm-raised-catfish-industry-2021-review-and-2022-outlook/
[2]: https://onlinelibrary.wiley.com/doi/epdf/10.1111/jwas.12619
[3]: https://iseralaska.org/static/legacy_publication_links/greatsalmonrun/SalmonReport_Ch_8.pdf
I'm really surprised by the low amount of suffering due to pork: intensive pig farming (crates etc) is pretty horrific and pigs are among the most sentient and intelligent meat animals, to the point that they are (or used to be) studied for stress response to help model humans.
Perhaps it's because so little pork is relatively eaten by Americans? Would you advise that people living in pork-heavy countries (some Asian, Eastern European etc) should also exclude pork if they want to follow ameliatarian diet?
Also, what about lamb? I'm assuming the amount of suffering is comparable or lower than beef, though the animals are much smaller, so how would that compare? And presumably wild game meat and "farmed" venison is lower on suffering than beef?
[I'm in the UK, if it makes a difference]
PS. It's probably of zero interest to you, but since reading your previous post on meat related suffering, I've excluded farmed salmon (I ate no other farmed fish) completely from my diet and it wasn't anywhere near as hard as I thought.
You seem to be using "Certified Humane" as an umbrella term, but it is not; rather, it is (FWICT) a legitimate standard that actually does provide for a reasonable standard of animal welfare.
The link you provide to the Humane Society's website has this to say about the Certified Humane label:
"The Certified Humane standards include minimum space allowances, bedding material and environmental enrichment (e.g., hay bales, pecking blocks and perches for chickens) among dozens of other basic requirements for animal health and nutrition. Animals are never confined in cages or crates and are free to display natural behavior. Farms are inspected by trained auditors."
This does sound like "functionally meaningless" to me. I have also previously looked into the Certified Humane standard in more detail, and also thought it seemed legit. I would certainly like to know if that is not the case.
https://certifiedhumane.org/
My understanding is that the US government standards all suck, but 3rd party standards (such as Certified Humane) may be legit. But (from my research) Canadian ones do not. I don't know about the UK, but I would like to.
That's interesting, the last time I read you on this topic you had written that while most egg labels are meaningless, that "pasture raised" was different and actually meaningful. Did the situation change or did your mind change?
Not sure if this is the case everywhere, but in many EU countries chickens' access to the outdoors is currently legally limited due to bird flu concerns.
Mmm- interesting, this.
My understanding is that virtually all eggs in the US are factory-farmed. In the UK (where I live), it is relatively easy to obtain "free-range" eggs (which is a regulated, protected term- it can't be applied by egg producers at will). Quite how strict the rules are, I'm not 100% sure, but the birds have to be allowed to walk around out of doors, and have access to water.
Chicken meat, however, is virtually all factored-farmed even in the UK. It is possible to find "free-range" chicken, but it's not easy and it costs 3-4 times the price. Any chicken-containing product or restaurant meal will almost certainly not use free-range chicken.
Are you able to summarize UK standards for me, or provide a good reference?
I think this defra document is probably fairly neutral, albeit long out of date (I don't believe standards have changed in the interim).
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/69366/pb6490-hens-020130.pdf
Two questions:
1. Farmed fish are often fed fishmeal, which is generally made from wild-caught fish. Since wild-caught fish are smaller and more numerous, plausibly the effects of eating farmed fish are dominated by effects on wild fish. How do you weigh this when making a specific recommendation about farmed fish?
2. If the reply to the above is that if the expected value on wild fish is unclear, since fish farming seems clearly bad, the expected value turns negative: It seems to me like a lot of demand for wild fish is “satisfied” by farmed fish (in that price increases in wild fish cause consumers to substitute to farmed fish if wild fish supply is relatively inelastic). So I wouldn’t be surprised if eating wild fish turns out to have very similar effects on increasing fish farming as eating farmed fish (perhaps with the exception of generally wild-caught species without obvious farmed substitutes, like anchovies). How different do you think the impacts on fish farming are from wild vs farmed fish consumption?
Thanks for the post!
Hi Tejas, I really appreciate your analysis of the second-order effects. I think about these issues too, so I tried to get some data on the problems.
I got estimates for salmon feed from [1] and picked the two most numerous fish in a salmon's diet with this feed. My estimate was about 75 anchovies and 55 sprats [2] So the second-order effects are roughly 2 orders of magnitude greater. The source also mentions farmed insects in the feed, which would further increase the second-order effects.
The blog post only seemed to imply that wild caught fish are killed before they would die naturally, making wild-caught fishing neutral. I agree that your mentioned effects might push this into negative territory. I have a lot of uncertainty around wild fishing for other reasons, but it could still be better than the counterfactual of chicken and shrimp. Chicken has its own bad second-order effects, and shrimp could dominate on scale (250+ shrimp = 1 salmon in calories).
Wild-caught fish has hovered around 90 million tons since 1990 while aquaculture has grown roughly exponentially to 120 million tons [3]. This fits the inelastic supply and increased fish farming story well. I'm not sure where the chain of effects bottoms out: demand for wild-caught -> more aquaculture -> increased demand for fishmeal/oil -> more demand for smaller fish… But it seems quite bad given that the substitution away from wild fish is likely toward farmed fish, shrimp, and chicken.
[1] https://www.skretting.com/siteassets/local-folders/norway/environmental-footprint-of-skretting-norway-salmon-feed-2018.pdf?v=48e941
[2] (1 anchovy / 0.002 kg anchovy oil)(25 kg anchovy oil / 100 kg fish oil)(0.1 kg fish oil / 1 kg feed)(1.2 kg feed / 1 kg salmon)(5 kg salmon) = 75 anchovies
(1 sprat / 0.001 kg sprat oil)(9.5 kg sprat oil / 100 kg fish oil)(0.1 kg fish oil / 1 kg feed)(1.2 kg feed / 1 kg salmon)(5 kg salmon) = 54 sprats
[3] https://ourworldindata.org/grapher/capture-fisheries-vs-aquaculture
Someone was inspired by this blog post to create a Tumblr: https://ameliatarianrecipes.tumblr.com
I really like this proposal!
There are two other levers that could be pulled to influence animal suffering caused by our diets:
1. Eating less meat and other animal products. The modern US diet contains far more animal protein than is optimal for our health. On average, we could eat a smaller amount of beef and cheese and come out ahead. When I've done things like this (for weight loss reasons), I found that there was actually pleasure in anticipating special meals and treats. "Oh, hey, tomorrow is cheeseburger day! It's going to be delicious!" Eating meat less often also allows for spending more money on the meat we do eat.
2. Only eating certain meat products where you have an extremely high degree of confidence that they were raised humanely. As you mentioned in the post, backyard eggs is a classic example of this. But plenty of people in rural areas have access to local farms, or to neighbors who, say, work together to raise a pig. In this case, it's feasible to actually inspect the farm! And this approach doesn't require being a coastal US elite with disposable income. Raising chickens for eggs is a classic low-tech, low-capital food source.
At the upper limit, if you really liked eggs, I suppose it would be possible to organize a consumer cooperative to raise chickens under genuinely humane conditions. Consumer cooperatives are awesome. They're halfway between a non-profit and a regular business, and they're democratically managed by their members. (Once they reach a certain scale, this usually means electing a board.) One big difference between consumer cooperatives and regular business is that consumer cooperatives usually don't scale beyond the needs of a local region, because they have limited access to capital. But if someone lives in the big city, and would enjoy the occasional omelet, it might work.
None of these options will provide the quantity of meat provided by factory farming, and they'll be more expensive. Which brings me back to (1), eating less meat.
This is a really good idea Ozy, I have basically been following this diet for years without thinking that I needed a word for it. Going to start using it and see if it catches on!
I like it and have adopted it. I was already moving this direction based on previous posts, but I like that it has a word.
Farmers' market eggs?
Shouldn't it be "amelitarianism"?
Some people eat mostly chicken and very little beef.
Other people eat mostly beef but very little chicken.
People who get most of their animal protein from chicken wouldn't take up this diet.
People who eat very little chicken anyway might but then it's not causing that much less suffering.
You are making the mistake of looking at the "average" and assuming that's "typical"
If only the chart you based it on wasn’t made up.
The asterisk link is broken.
No restrictions on hunted animals like deer, I assume?