[ACC] Is Eating Meat A Net Harm?

[This is an entry to the 2019 Adversarial Collaboration Contest by David G and Froolow. Please also note my correction to yesterday’s entry.]


Introduction

Many people around the world have strong convictions about eating animals. These are often based on vague intuitions which results in unproductive swapping of opinions between vegetarians and meat eaters. The goal of this collaboration is to investigate all relevant considerations from a shared frame of reference.

To help ground this discussion we have produced a decision aid making explicit everything discussed below. You can download it here and we encourage you to play around with it.

The central question is whether factory farmed animal lives are worth living; the realistic alternative to meat eating is not a better life but for those animals to not exist in the first place.

We begin by investigating which animals are conscious. Then, we compare the happiness literature to the conditions under which animals are factory farmed to figure out if from their perspective non-existence is preferable. And finally, we survey the more easily measurable impacts of meat eating on environment, finance, and health.

1. Consciousness

1.1. What is consciousness?

This essay isn’t about a general theory of consciousness. We tried to research this and our main takeaway is simply that consciousness is really, really weird. When we say that something is ‘conscious’ we mean simply that it’s ‘like something to be that thing,’ and that if we were that thing we’d care if ‘good’ or ‘bad’ things happened to us.

Of particular relevance is the conscious experience of pain and suffering, which we regard as morally undesirable when it occurs in ourselves or others. Most animals have damage sensors, but triggering these may not result in the subjective experience of ‘suffering’ if the animal is not conscious or the stimulus is a constant presence that has been accustomed to.

And this is the absolute crux of this investigation; if animals suffer under current farming standards to the point of preferring non-existence then there is a moral burden on meat eaters to justify eating them.

To resolve this, we will look for in different animals at two kinds of evidence for consciousness:

1. A brain architecture similar to humans resulting from the same evolutionary process
2. Behaviors that are hard to explain except with reference to having experiences

1.2. What parts of the brain are ‘responsible’ for consciousness?

You might assume consciousness is just caused by ‘the brain,’ because without it how could we think (or do) anything? However, a huge part of intelligent information processing happens in our brain without giving rise to conscious experience.

Try noticing what happens when you read the sentence “The dustmen said they would refuse to collect the refuse without a raise.” Notice how the word ‘the’ appeared twice? And how you read it both times despite Scott’s best efforts at conditioning you otherwise? Very good. Now, notice also that the word “refuse” appeared twice, but the first time you not only interpreted it but heard in your mind as a verb with the stress on the second syllable, and the second time as a noun. Before the words appeared in your conscious mind, as visual on a screen, as sound in your head, as a feeling of understanding the sentence, your brain already did all the hard parts of figuring out what they mean subconsciously, without you experiencing anything.

If you sever the spinal cord, the 1 billion nerve cells in the backbone connecting your brain to your limbs, you would lose complete sensation below the neck but you would continue to have thoughts and rich experiences. Or consider the cerebellum, the 69 billion (80%+) neurons which are responsible for motor control. If you cut off parts of it you may lose the mechanic ability to execute certain motions like playing a piano or walking but you will remain completely conscious with a sense of self, a memory, and the ability to plan for the future.

What is in common between the spinal cord and the cerebellum? The neural circuits operate in parallel with hundreds of independent input/output logic gates and all fire one way, instead of forming an interconnected multi-way circuit. There is no capacity for reflection, just predefined decision making. A necessary condition for consciousness (in people) is two neurons ‘talking’ to each other, rather than just passing information with deterministic modification up the chain of command.

Where consciousness appears to be generated is the posterior cerebral cortex, the outer surface of the brain composed of two highly folded sheets. Stimulate it electromagnetically and it’s like any other acid trip. The folded physical structure seems extremely important for generating consciousness because it maximizes the surface area exposure of neurons to each other. Many parts of your brain can be removed without major changes to your personality of intelligence, but if even small parts of the posterior cortex are missing surgery patients lose entire classes of conscious content: awareness of motion, space, sounds, etc.

It’s important to recognize that consciousness is not simply ‘caused by anything that happens in your brain.’ It’s a specific, fragile thing with distinct characteristics that differ from other neural activity that we associate with intelligence – therefore the relative intelligence of animals to humans does not necessarily map closely to their degree of consciousness.

1.3. Neural indicators of consciousness

All mammals have a cerebral cortex. Mice and rats have a smooth one; cats and dogs have some folding; and humans/dolphins/elephants have highly expanded and folded cortices. Therefore all mammals are probably conscious, although with large differences in vividness and complexity.

Birds and reptiles are a harder case because their brain evolution diverged much earlier. They have instead a cluster of neurons with chemical markers associated with differentiated layers of the neocortex but without the folded shape that maximizes connectivity.

By contrast, fish do not have any neural architecture unique to the consciousness-related parts of the brain and are probably unable to feel fear or pain in the way a human would – we strongly encourage you to read this article in full to convince yourself of this claim. Although fish show pain-like responses to harmful stimulus and do so less if given painkillers, this is true even when the entire telencephalon (which includes the forebrain) is removed so on balance it is unlikely they are having a qualitative experience to accompany that response.

1.4. Behavioral indicators of consciousness

Behavior seems like an obvious place to look for evidence of consciousness. However, any behavior can be explained by intelligence alone, or even sub-intelligent evolutionary ‘hard coding’. If you swat a fly, it will make loud ‘angry’ noises and go away. Your little brother would react the same. If you knew nothing about the neural architecture of flies, you might conclude that flies are just as conscious and capable of suffering as people.

One way around this is if we can design tests that indirectly look for mental states, such as the mirror test (whether an animal can recognize its reflection). But elephants (definitely conscious) routinely fail and at least one fish has passed, so we are wary about assigning much weight to these tests.

Another is to look for behaviors that map onto extremely complex emotional states that we observe in humans. If there is a large difference in intelligence but a great similarity in behavior, we can infer that the animal is having a similar conscious experience.

Starting simple; if you play with a dog it will act in the highly specific ways you might if you were feeling ‘joy.’ From a hormonal and intelligence perspective, stress and positive excitement are very similar states, and in non-conscious creatures we would have no reason to expect – for example – a creature to seek out stressors like a chew toy unless they had some positive feelings towards them. That we can so clearly tell how a dog is feeling is to us highly persuasive evidence of consciousness.

Dogs also exhibit something quite analogous to a theory of mind, for example they will comfort their owner if their owner is sad (but maybe this is a learned behavior)

Dogs are unlikely to be a special case; other animals of varying intelligence also exhibit complex behaviors indicative of consciousness:

• Chimps who see another chimp lose a fight will direct more grooming behavior towards the loser, but not if they don’t see that chimp lose the fight. [Link, popular coverage]
• Corvids who hide a treat when being observed will sneak back later and rehide the treat somewhere else, indicating (perhaps) a theory of mind and (certainly) ‘mental time travel’ of imagining the self in various future states. [Link, popular coverage]
• Dolphins given a test to discriminate between X and Y for a reward, but including the option of ‘bailing out’ of the test in exchange for a lesser reward, will bail out more often in more difficult tests, indicating a theory of metacognition (which we’d say is adjacent to – if not the same thing as – a theory of mind). [Link, popular coverage]

Spending time with animals (higher mammals, especially) makes it extremely hard to imagine they are anything but conscious, but we recognize that any behavior could be explained as an expression of intelligence without assuming conscious experience. However, we are reasonably confident that:

1. The range and complexity of behaviors conducted by animals correlates closely with the brain architecture we believe causes consciousness – the more complex the brain architecture, the more consciousness-like the behavior. This would be a substantial coincidence if in fact animals were not conscious.
2. Animals we intuit as conscious are less likely to exhibit ‘glitching’ behavior indicative of being a non-conscious rule-following automaton. There are many examples of ‘glitches’ in insect behavior (such as ant vortexes of death, repetitive digger wasp behavior [although maybe not] and moths failing to notice they are circling a candle), whereas there are very few examples of ‘glitches’ in mammal behavior. A humorous example of a glitch in bird behavior can be found in YouTube videos where the ‘imprinting mechanism’ of ducklings has confused them into thinking a dog is their mother.

1.5. What animals are conscious?

It’s fair to reflect on the uncertainty in the above, but we’d be comfortable ascribing consciousness on the basis of neural architecture and behavior as follows:

There is good reason to believe all common land-based food mammals (cows, pigs, sheep, goats) are highly consciousness. On the other side, we think we can be reasonably confident fish don’t suffer in a morally relevant way. We’re not sure about chickens. We encourage you to read this overview of their behavior in full to convince yourself that their emotional and cognitive intelligence would group them with simple mammals if they had the same neural architecture.

However, since in most parts of the human brain ‘intelligence’ does not correspond to ‘consciousness,’ and because chicken brains are a clump of neurons with a different evolutionary history and lacking the distinct layered and highly folded structure of the cerebrum, in the model we assume their likelihood of consciousness is 75%.

A key part of this post is to quantify vague feelings about animal consciousness. This is similar to what Scott did with a sample of Tumblr respondents here and SSC reader Tibbar did with an MTurk sample here. Their results are expressed in terms of an animal’s ‘worth’ relative to a human in percentage terms.

% Consciousness

Tumblr sample

MTurk sample

Human

100

100

Chimp

20

50

Elephant

14

100 (!)

Pig

3

20

Cow

2

33

Chicken

0.2

4

Lobster

0.03

1.6

However, it’s important to make the distinction between ‘worth of experience’ and ‘worth of suffering’ because while we might rather be a human than a chicken on a good day, feeling pain might be equally unpleasant in either body. Below is our best guess for a ‘universal’ estimate (i.e. even a meat eater ought to agree that these are plausible) – people who place a premium on animal experience, such as many vegetarians, would likely rate animal experiences higher:

%Weight Suffering

%Weight Experience

Human

100

100

Chimp

90

50

Elephant

90

35

Pig

80

25

Cow

50

10

Chicken

10

1

Lobster

0.1

~0

What we mean by the above is, for example, if the unit of suffering was ‘being boiled alive,’ based on our understanding of how vivid their sense data would be assuming consciousness we would be roughly indifferent between being boiled once as any of a human, chimp, or elephant, 10 times as a chicken, or 1000 times as a lobster. However, we would be indifferent between living (assuming no scarcity or predators) for 1 year as a human, 2 as a chimp, 4 as a pig, or 100 as a chicken.

In the model, the moral impact of a farmed animal is its likelihood of consciousness times the moral weight of its suffering if its life is, based on information in the following sections, ‘worse than non-existence,’ otherwise the moral weight of its existence.

2. How many animals are farmed and under what conditions?

2.1. Animals eaten per capita

The OECD records the exact weight of meat consumed per year, and so by dividing this by the carcass weight we get the per-year per-capita animal consumption by country. This is consistent with other estimates on the web.

Animals / Capita consumed by four major categories in 2018

Australia

Canada

EU27

UK

USA

World

BEEF

0.05

0.05

0.03

0.03

0.07

0.02

PIG

0.27

0.20

0.44

0.22

0.29

0.15

POULTRY

29.42

22.53

15.70

18.53

33.12

9.44

SHEEP

0.38

0.05

0.07

0.22

0.02

0.09

The conditions most animals are farmed in may surprise you. Since a few large factory farms account for most animals farmed, the typical farm may well be a small mom-and-pop operation but the typical food animal is raised on an industrial scale. Further, ‘ag gag’ laws limit facts about animal conditions reaching public awareness and are arguably designed explicitly to allow companies to mislead consumers about the conditions most animals are farmed in.

Animal rights organizations will frequently quote a study by the Sentience Institute that over 99% of animals eaten in the US are factory farmed. Although a biased source, it’s consistent with other government estimates and we get similar results when replicating their methodology with USDA figures. What really drives this statistic is, as seen above, chickens form the vast majority of the farmed animal count and are almost exclusively farmed industrially.

 

My
calculation for % factory farmed using Sentience Institute
methodology

Sentience
Institute estimate for % factory farmed

BEEF

64%

70%

PIG

94%

98%

POULTRY

~100%

~100%

SHEEP

64%
(based on being woolly cows)

Not
included

Comparing across countries is difficult, but it seems that America is slightly more industrialized than the EU. My best estimate is that the difference is not significant enough to make a moral difference. If you eat meat and cannot explicitly trace the source you are most likely eating factory farmed meat.

2.2. What is it like to be farmed industrially?

The definitive feature of factory farming is that market incentives lead to a paperclip maximizer situation where producing as many animals as possible takes precedent over concerns about animal welfare. Consequently, the ‘experience’ of being factory farmed is best understood as a particular form of slavery where cruelty is the side effect of a system designed to maximize economic output.

2.2.1. Chickens

Chickens can be raised in two ways: in cages or in a shed. Cage-rearing chickens is typical in developing economies such as China, but in the West cages are used for egg laying hens only. Slaughter chickens in the West are raised instead in a large ‘broiler’ shed covered with liter.

Broadly speaking, caged chickens have literally no human analogue in terms of how much they suffer. They live in a state of constant pain and anxiety, barely able to move. The only mercy is that they do not suffer for long. In this analysis we focus on meat-eating in a Western context. So, we model 100% of chickens as being broiler farmed.

A factory farmed slaughter chicken lives for approximately 47 days, during which time it grows to a weight of 2.6kg (42 days and 2.5kg in the EU). This is analogous to a newborn human baby reaching adult weight by the first birthday. To achieve this rapid weight onset, a combination of force-feeding, drugs and high-energy feed is used. But the worst culprit is selective breeding. In a study by Kestin, between 2% and 30% of broiler chickens, depending on the breed, had a gait score between 3 and 5 on a 5 point scale (1 no issues, 3 obvious gate defect, 5 unable to move at all). But 100% of a control group bred randomly and then raised under the same conditions had no or minor mobility issues. Selection for quick growth rather than fitness in the wild leads to a high rate of heart attacks and other organ failure. In the final weeks of life, the chickens often outgrow the ability of their legs to support them, making broken or otherwise failed legs endemic in the industry.

Photos of cage-raised chickens, borderline NSFW

Photos of broiler shed chickens, NSFW

Because it is cheaper to only change the litter between flocks it is a major source of bacterial infection and especially contact dermatitis (rashes and lesions on the chicken’s feet and lower body). It is common practice in the EU (not the US) to remove a portion of the flock a week before slaughter time to create enough space for the remaining birds to reach their usual slaughter weight, suggesting there isn’t much free space for the birds. Birds whose legs fail will often dehydrate to death. We don’t want to overegg this – a dead bird is an unproductive bird and only around 3.3% of the flock die during growth for any reason – but remember that this is a 3.3% chance of dying in only six or seven weeks.

De-beaking is common in broiler chickens (universal in laying chickens). One reason for debeaking is to reduce cannibalism which occurs because the birds are so stressed – pet chickens will peck each other to establish a dominance hierarchy but don’t kill and eat each other. Beaks are sensory and manipulative tools for chickens, so this is analogous to cutting the fingers of prisoners off without anesthetic to lower the probability of escape.

Photos of debeaking, NSFW

Shed chickens have it slightly better. They have a small amount of mobility, are able to do some natural activities such as socializing and digging in the dirt with their claws (but not usually their beaks) and have a little natural light from windows in the warehouse. On the other hand, chickens only cannibalize each other when very stressed and the strain on their systems from the massive growth they are forced to undertake causes considerable pain.

We think it is reasonable to say that broiler chickens exist in a state worse than death – in the model, we assume chicken-days are equivalent to -2 human-days (you’d rather have your life be 2 days shorter than have to experience a day of chicken life), but your intuitions may differ substantially.

2.2.2. Pigs

Pigs are the next-most commonly farmed food animal. There are two major sources of cruelty in pig production; the raising of the food-pigs themselves and the creation of new food-pigs from breeding pigs.

Breeding sows are confined to a ‘gestation’ or ‘sow’ crate for most of their lives. These are only slightly larger than their bodies, making it impossible to turn around or even lie down. Generally the floors are made of slats or iron rungs to allow manure to fall through. These slats can hurt the sensitive feet of the pigs, and the fact that they are confined directly above their own manure means they are exposed to ammonia toxicity, which leads to respiratory conditions common in confined sows (and presumably smells incredibly distressing). Pigs are highly intelligent, and the unstimulating confinement means that the pigs engage in repetitive stress behaviors such as biting at the metal bars of their cage – this can cause further harm such as mouth sores.

Shortly before birth, the pregnant sow is taken to a ‘farrowing crate’ – even more restrictive than a sow crate. This is designed to separate the mother from the piglet so the piglet can nurse without being crushed (piglets being crushed can happen in the wild, but it is rare – this is a problem almost entirely caused by the confined conditions the sow is kept in). The crate is so tight the mother cannot even see her baby once it is born, and the baby is taken away after about 17-20 days. The piglet is then prepared to be fattened for slaughter, and the mother is either re-impregnated and returned to the gestation crate or slaughtered herself if she is unlikely to survive another pregnancy.

Photos of gestation and farrowing crates, surprisingly SFW

Piglets being prepared for slaughter are castrated and have their tails docked, often without anesthetic. Unlike chicken beaks, pig tails don’t really seem to serve any purpose, but pigs show pain behavior towards their stumps suggesting that it is very sensitive even after being docked. The tails are docked to prevent other pigs biting it and causing an infection – again, behavior which is vanishingly rare in the wild and therefore seems to be a stress response to the conditions they are kept in. Piglets may also have their teeth clipped to prevent biting but we can’t find figures on how common it is. Pigs prepared for slaughter are kept in ‘finishing crates’ which seem to run from anywhere between a slightly larger sow crate (larger only in the sense that it is bigger – finishing pigs are much larger so don’t have any more space to turn around or express natural behaviors) and something a little more like a traditional farmyard pen but indoors – six or seven pigs confined to a small pen where they have just enough space to walk around if they want to.

Pig-tures of finishing crates, SFW

Pigs are highly intelligent animals, and when not confined to stalls will spend hours playing and rooting around in the mud. The pigs consumed for food will be in constant low-level pain and sows used for breeding will be in quite intense pain constantly. It is hard to imagine a more distressing event than having your child taken away from you or being taken away from your mother, and we might imagine that the constant lack of stimulus for both food and breeding pigs causes considerable boredom and sadness.

It is a harder call whether pigs exist in conditions worse than death. My intuition is that food pigs are right on the border, and breeding pigs would strongly prefer to not live. In the model we assume that a pig-day is worth -1 human days.

2.2.3. Cows

Cows are the only animals routinely farmed in conditions approaching the way people imagine farming to be in their head – that is, in a field where they have enough space to move around and socialize. Factory farmed cows spend six to twelve months being raised outdoors in fields (the sorts of cows you see dotted around the countryside), and are then transported to ‘feedlots’ for their last few months where they are fed an artificial diet of corn and soy that is very hard on their bodies and can cause illnesses such as ulcers. Note that almost all cow-meat can be labelled ‘grass fed’ because most cows spend their first year in fields eating grass: doesn’t mean that is where the majority of their final slaughter weight came from! Much like pigs, cows have complex social hierarchies and being put on a feedlot with thousands of other cows is depressing.

Pictures of feedlot, SFW

Beef cattle also endure individual painful events like castration, branding and dehorning (often done without anesthetic) and transport in cramped crates for long periods of time. It is unclear how to incorporate the impact of these events on the animal’s overall quality of life. We’d suggest animals raised on a field have the same quality of life as a traditionally farmed animal and animals raised on a feedlot have a quality of life moderately worse than a typical elderly human. By quite a long way cows appear to have it the best of all factory farmed animals and have lives that are clearly worth living.

We think we’d be pretty content to live as a cow in a field, but cows on feedlots seem to have lives that are closer to food-pigs. Approximately averaging these out over a cows’ lifetime, we model 1 human day as equal to 10 cow days. We can’t find good information on how sheep are factory farmed so we’ve assumed they’re just woolly cows for the purpose of estimating their quality of life.

2.2.4. Value of animal life

Quantifying and comparing subjective experiences of farmed animals is hard because there is no ‘natural unit’ of suffering or experience. We proceed instead by asking ‘at what factor are we willing to skip/trade off days of my life to live as any particular animal under factory farmed conditions’ as described above. Factory farmed cow lives seem greatly preferable to non-existence, pigs spend most of their time experiencing some form of chronic or acute suffering, and chickens have truly awful lives from birth to slaughter.

Slaughter itself may be a morally relevant part of valuing an animal’s life. In principle, animals are stunned before slaughter so that the process is painless (and evidence suggests animals are not distressed by watching other animals being stunned). However, in practice animals are often not insensible to pain when they are skinned or carved up, either because of poor training (paywalled WaPost link), religious beliefs around the way meat should be prepared (link, further figures) or just because of a culture of laxity and cruelty (INTENSELY NSFW link of animals being abused by slaughterhouse workers, SFW-ish PDF report of the investigation). ‘Ag gag’ laws and other efforts by farmers to avoid bad press prevent serious scholarly investigation of the extent of the issue, but the AnimalAid hidden camera investigation linked as a PDF above found evidence of criminally cruel treatment at one of the three abattoirs observed and evidence of mistreatment at another.

Also, some people believe in a specific ethical obligation to not kill conscious creatures that do not want to be dead, so that slaughter of even humanely stunned animals is immoral. We instead take the consequentialist view that there is a symmetric value in actualizing the existence of conscious creatures that want to be alive (again, a farmed animal’s practical alternative is non-existence).

In the model, we assume that at human-level consciousness the experience of a typical human life-day is worth the factory farmed experience of 10 cow days, -1 pig days, and -0.5 chicken days. We value a year of perfect health at $50,000 in line with typical healthcare priority setting in the US (it is much less in the UK and Europe – closer to £30,000), where a typical western life-year is about 86% of that ($43,000). Slaughter is not included as even the cruelest slaughter imaginable would be QALY-negligible if averaged over an animal’s whole lifespan.

3. What lives are worth living?

Speaking loosely, evolution does not care how happy your life is as long as you a) exist and b) pass on your genes, and so has come up with a number of ‘patches’ to the conscious reward system to ensure animals are never too satisfied to stop competing to breed but never so dissatisfied they would prefer to be dead. Instead, what we have is roughly a baseline set of happiness from which we deviate when good or bad things happen, but to which we almost always return to. If this is also true of animals then it does not matter that we perceive their lives as described above as intolerable; if we were actually forced to live as that animal then we would observe ourselves hoping we don’t die painlessly in our sleep. Since we cannot ask animals directly if they consider their lives worth living, we instead look at the conditions in which people report changes in happiness or commit suicide, and compare these to the lived experience of factory farm animals.

3.1. Habituation and Happiness Set Points

Habituation is a “decrease in response to a stimulus after repeated presentations.” The simplest form of learning, it is caused by neural processes that regulate responsiveness to different stimuli. When we are repeatedly sent a signal, especially if it is highly frequent and hasn’t recently changed in intensity or duration, we consciously experience it less acutely. This makes sense. If consciousness is about complex reflection, after we’ve already processed a signal and determined a response if any, the response becomes subconsciously automatic.

However, we don’t habituate just to local physical sensations, like the ticking of a clock or the pressure of a shirt against our skin. We habituate to pain and suffering as well, even to large shocks to the system. In the literature this is known as the disability paradox, whereby a majority of those with severe disabilities report having a good or decent quality of life, even when to external observers it seems like a life not worth living (although this story is nuanced, and some of the improvement is related to the ability of intelligent humans to adapt by changing their lifestyle).

Nevertheless, the consensus in happiness research is that people have a fairly stable general level of baseline happiness which they return to after certain large changes. In a famous study by Brickman (1978), paraplegics and lottery winners reported similar levels of happiness before and after what one might assume was a life-altering development, either extremely negative or extremely positive. And in a twin study of several thousand by Lykken and Tellegen, it was found that about 50% of variation in Well-Being scale of the Multidimensional Personality Questionnaire is associated with genetic variation, and less than 3% (!) of the variance can be accounted for by any of socioeconomic status, educational attainment, family income, marital status, or religious commitment.

3.2. Lessons from suicide

Sometimes, humans decide that their lives are intolerable and commit suicide. Interestingly, we basically never observe this in other animals. The only maybe-credible anecdotal claims are of dolphins, which are highly intelligent and can commit suicide by not breathing (breathing by dolphins is probably an active choice, rather than an automatic process as regulated in humans by the amygdala).

This doesn’t mean we can conclude farmed animals prefer living because animals might lack the theory of mind or intelligence to act on their preference to stop existing. However, if humans in extremely poor conditions overwhelmingly do not choose suicide, we might infer that animal lives of roughly similar quality would also be worth living. Two well-studied areas where humans are placed in extremely poor conditions are slavery and terminal disease.

3.2.1. Slavery

The historical consensus is that while slavery caused extreme stress and suffering, the rate of suicide by black slaves was quite low. According to the 1850 U.S. census, slaves had a suicide rate of 0.72 per 100,000 while whites had a rate of 2.37 and freed slaves a rate of 1.15. From the Federal Writer’s Project Slave Narratives which documented incidences of resistance, only 1.2% were acts of suicide. Further, when slaves did resort to suicide, it was usually in response to deterioration in their circumstances or unfulfilled expectations, rather than being explained by living under the most brutal conditions – this is consistent with a ‘happiness set point’ theory.

To be clear, we are not saying that because enslaved Africans committed suicide at lower rates than free whites slavery wasn’t ‘that bad.’ There is a substantial academic literature explaining the cultural reasons for the difference. The observation is simply that the main explanatory factor of whether a slave thought it worth living was not how bad his or her life objectively was.

3.2.2. Terminal Patients

In a review of the psychological profile of patients in palliative care of 18,000 terminal cancer patients, a small number of which committed suicide, it was found that some of those who committed suicide:

“…presented functional and physical impairments, uncontrolled pain, awareness of being in the terminal stage, and mild to moderate depression… however, the loss of, and the fear of losing, autonomy and their independence and of being a burden on others were the most relevant.”

The presence of significant pain or even depression (what we might refer to as ‘objective suffering’) was not a significant factor in predicting suicide, the best revealed preference we have for whether a life is considered worth living by the morally relevant actor experiencing it.

3.3. What does this mean?

Preference for living is a strongly mean reverting process. The scientific literature and historical examples from slavery and terminal illness both suggest that humans will habituate to almost anything that is done to them. In our ancestral environment life was really, really hard. Brutally hard. And it makes sense that even in environments that modern folks would instantly label as ‘much worse than non-existence,’ evolution made sure that we would continue to have the strength of will to not only survive but want to.

How far does this go? Does this mean that animals, no matter how much suffering they experience, prefer living? Reasonable people can disagree. As detailed above, factory farmed animals – especially chickens – do not exist in anything remotely resembling an ‘ancestral environment’. Chronic stressors are more likely to cause permanent changes in happiness than acute changes, and, as in the case of chicken cannibalism absent food scarcity, it is likely that factory farming creates an environment that is not only unpleasant but one which even the astounding level of habituation we observe in humans might not adjust above the ‘worth living’ watermark.

In the model we take the habituation literature seriously, estimating by how much deviations from long-term set point of average quality of life (0.86) are neutralized by habituation. Based on the previous sections and how much worse factory farmed conditions are than an ‘ancestral environment’ in which habituation would be calibrated, we estimate that humans habituate by 80%, cows 70%, pigs 60%, and chickens 50%. Even though before we had assumed that we would prefer non-existence to being a factory farmed pig, and would dislike being a chicken twice as much, once habituation is taken into account pig lives are slightly better than non-existence, and chicken lives are quite bad but have a (negative) moral weight not much larger than the positive one of cows. However, since we consume so many chickens in the end they dominate the analysis.

4. Health Considerations

4.1. Nutrition

Animal protein sources such as meat, fish, eggs, and dairy contain a good balance of the 20 amino acids that we need for almost every metabolic process in the body, whereas individual plants are generally deficient in mix or concentration. The same is the case for micronutrients: animal protein sources are much higher in vitamin B12, vitamin D, the omega-3 fatty acid DHA, heme-iron and zinc.

Animal products provide most of the zinc in US diets, and meat, poultry, and fish provide iron in the highly bioavailable heme form. For example, the panel setting the new Dietary Reference Intakes recommends an 80% higher daily iron intake for vegetarians.

Concern also has been expressed about the difficulty that children have in obtaining adequate energy and nutrient intake from bulky plant-based diets. Dutch infants consuming vegan diets had poorer nutritional status and were more likely to have rickets and deficiencies of vitamin B-12 and iron, and the World Health Organization strongly recommends animal products for infants to ensure enough calcium, iron and zinc.

Whether or not a plant based diet is viable from a nutritional perspective depends mostly on whether you have the economic means to consume a wide variety of food sources, and may be riskier for small children or those whose ancestry is from regions where meat-eating was prevalent.

4.2. Long Term Health Outcomes

Estimating the long-term health outcomes of eating certain things is difficult because food is highly bound up in the culture we live in and culture correlates to just about every health outcome you could possibly imagine. Even less conveniently, nutritional science is highly anti-inductive; if a particular food group is identified as being healthy people with an interest in being healthy flock to that food group, and people with an interest in being healthy are likely to be healthy for a bunch of reasons regardless of diet.

So here’s a nice headline result: vegetarians have less heart disease with extremely high certainty, and probably less cardiovascular disease and cancer too. Most of the studies in that meta-analysis have had some of the really obvious stuff adjusted away (race, income, etc.) but not all studies adjust for all confounders, and we should be cautious about trusting studies that ‘adjust for confounders’. If you ignore confounders then the answer is clear; eating vegetarian is good for you in every single way we can measure (including, possibly, circulating testosterone in defiance of stereotypes about meat eaters!).

If you are interested in confounders: There are a handful of cool natural experiments, taking groups with reasons to eat certain food but not bother with the associated healthy lifestyles, which are the closest we are likely to come to a true experiment in this area. In particular, the American Adventist Health Studies are pretty much state of the art in the field from what we can see. Adventists have quite unique dietary habits, brought about by religious prohibitions on certain foodstuffs which some Adventist churches follow and some don’t. Consequently, if you are an Adventist you are functionally ‘randomized’ into different food-eating conditions depending on which church you attend, and this randomization can be exploited by researchers.

Based on the Adventist Health Studies, a vegetarian diet increases life expectancy by around 3.6 years. The less meat you eat, the healthier your BMI and the less likely you are to get diabetes.

Overall we might expect lacto-ovo vegetarians to have a health related quality of life around 10% better than a meat eater, with most of this benefit being apparent 20 years after making the switch to a vegetarian diet.

You could complicate this picture a lot (especially by introducing future discounting) but we think the general principle that if you value life-years towards the end of your life you should likely go vegetarian is well demonstrated by the data:

One final point on how meat might affect your lifespan; there is a growing awareness of the fact that industrially produced meat is an ideal breeding ground for zoonotic disease, and that those diseases can mutate and jump to humans very quickly. Previous pandemics such as H1N1 (‘swine flu’) and H5N1 (‘bird flu’) may have originated with farmed animals, and were rapidly spread by the close contact of unhealthy animals and global nature of the meat supply chain. At the margin, eating meat probably increases the probability of a global pandemic but there isn’t good evidence on how much your individual consumption affects things at the margin.

In the model we take the Adventist study result at almost face value, estimating that eating vegetarian will increase your lifespan by 3 years, and include constant low costs due to possible nutritional deficiency and moderate benefits to health that appear later in life.

5. Environmental Impact

The environmental impact of meat consumption is difficult to measure and aggregate because the numbers are sensitive to the type of farming used and location, and any serious attempt requires a massive aggregation of different data sources. The best we could find was by Oxford’s Zoology department which combined data from 570 studies with a median reference year of 2010 covering 40,000 farms and 1600 processors, packaging types, and retailers, in 119 countries of 40 products representing ~90% of global protein and calorie use, focusing on five environmental impact indicators: land use, freshwater withdrawals weighed by local scarcity, greenhouse gasses, acidifying emissions, and eutrophying emissions.

Overall, today’s food supply chain creates ~13.8 billion metric tons of CO2 equivalents, which is about 26% of human caused emissions. It also causes 32% of global terrestrial acidification and 78% of eutrophication. It’s also very resource intensive, covering about 40% of the world’s ice- and desert- free land, and driving roughly 90% of global scarcity-weighted water use because irrigation returns less water to rivers and groundwater than industrial and municipal uses and predominates in water-scare areas and times of the year.

Because of different technologies and other environmental variables, the environmental impact of any foodstuff can vary widely. For example, ninetieth-percentile GHG emissions of beef are 105kg of CO2eq per 100g of protein, and land use (area multiplied by years occupied) is 370 m2 ∙year. These values are 12 and 50 times greater than 10th-percentile dairy beef impacts.

However, as you can see below the environmental impact from meat dwarfs that of other nutrition sources:

In total, meat, aquaculture, eggs, and dairy use ~83% of the world’s farmland and contribute 56-58% of food’s different emissions, despite providing only 37% of our protein and 18% of our calories.

Because of substitution effects and nutritional requirements, it is unclear exactly how much of these resources would be freed up if we switched away from eating meat. In a simple model where we assume ‘protein is protein and calories are calories’ and freed up land would only remove carbon through natural vegetation and accumulated soil carbon, “moving from current diets to a diet that excludes animal products would result in reducing food’s land use by 3.1 (2.8-3.3) billion hectares (a 76% reduction), including a 19% reduction in arable land; food’s GHG emissions by 6.6 (5.5-7.4) billion metric tons of CO2eq (a 49% reduction); acidification by 50% (45-54%); 19 eutrophication by 49% (37-56%); and scarcity-weighted freshwater withdrawals by 19% (−5 to 32%) for a 2010 reference year.”

It’s difficult to translate these tradeoffs into ‘one number’ that captures the environmental impact of meat eating. Land which currently supports animal farming is likeliest to be least suitable for agriculture or urban dwellings, and the costs of climate change, the value of species diversity, and the future scarcity of freshwater are all difficult to measure.
One (very approximate) way is to assume that the human race is polluting the planet as much as it can sustainably (at current technologies it’s a lot worse, but we’re science-optimists), and that the long-term impact of a reduction in pollution is a proportionally inverse change in the equilibrium human population.

Water shortages, eutrophication, and acidifications are serious environmental concerns but can be managed. Greenhouse gas emissions and land seem like the most important constraints. Combining the statistics from the total absolute resource impact of the food supply chain with the relative impact of switching from meat-eating, we get that if the planet went vegetarian we’d reduce emissions by 12.5% and free up 30% of the world’s non-desert/ice land most of which we would not be able to immediately put to good use. We think it’s reasonable that the total reduction in ‘human pollution and resource use’ would be about 10%. Since raw resources and access to clean environment aren’t the only limiting factors on population size this result should be adjusted downward by a reasonable factor.

In the model, we assume that without meat farming there would be about 2.5% more capacity for population or quality of life, 30% of which (completely uneducated guess) would actualize as more lives, and 70% of which would actualize as better lives (and would count at 1/5th weight due to habituation).

6. Cost of Switching Diets

One good reason not to switch to a vegetarian diet would be if doing so was prohibitive, either because of the financial cost or the satisfaction from eating.

6.1. Cost per meal

In a trivial sense, vegetarianism is clearly cheaper. It takes more time and energy to grow plants that we feed to animals and then eat the animals than it does to just eat the plants themselves. This is borne out by research into the cost per calorie of various foodstuffs. Of course, humans don’t eat exactly the same food animals eat, and vegetarians are for some reason unwilling to just drink 2000 calories of canola oil every day.

The cost of various types of diet seem to be bizarrely under-studied (or perhaps crowded out by the literature on trying to get people to stop eating junk food). The one academic source I found seems to be really high quality though. Here is the paper and here is a nice associated blog post.

At all income levels meat eaters spend about $20 more per week than true vegetarians (~$1000 / year). Adjusting for all controls (including politics and body weight which may be affected by vegetarianism) reduces this number to a savings of $11.1/week, which is what we use in the model.

6.2. Psychological costs

However, one switching cost that might not be trivial is the psychological importance of having meat in your diet.
Most vegetarians eventually enjoy vegetarian food as much as meat (not sure if that’s just survivorship bias), but anecdotal experience from everyone I know who has gone vegetarian says that there is a really horrible period of adjustment of at least a couple years where you want to eat meat and can’t.

One reasonable measure of psychological pain is to look at how much people would be willing to pay to avoid it, which conveniently has been studied:

This table is the output of a point regression asking US consumers how much they would pay to avoid a one percent decrease in each category of food. A 1993 dollar is approximately half of a 2019 dollar (1.78), so consumers are saying that they would pay $15 per year to avoid a one percent reduction in their meat consumption. It is highly unlikely that people would accept one hundred times this value to cut out meat completely – it’s easy to cut out the first few percentage points of meat (just have slightly smaller portions) but gets harder as you are forced to make fundamental changes to your diet.

For those inclined to stop eating meat, we wouldn’t overthink this parameter. If the habituation literature has convinced you that you’d be just as happy in a wheelchair as a lottery winner then in the long-run you probably won’t mind eating more tofu.

In the model, we double the $1,500/year preference loss of meat implied by marginal preferences to $3,000 to account for social costs and elasticity of demand, and assume this decays to 0 by 10%/year as one gets used to the new lifestyle.

7. Conclusion

Overall, the case for reduced meat consumption is strong. Vegetarianism is cheaper, better for your health (if you can afford a diverse diet and are not an infant), and is less impactful for the environment. It also has a significant moral cost in terms of animal suffering.

At the outset of the collaboration the vegetarian was sure that farmed animals’ lives are so awful that the status quo is an unmitigated moral disaster; the meat eater was open to that conclusion but could also imagine being persuaded to spend all disposable income on buying meat and throwing it away because that was the only efficient means of causing the existence of sentient creatures who strongly prefer to exist. If you mapped reasonable conclusions on meat eating from a scale of -10 to 10, you could say that we started out as a confident -5 and highly uncertain 2, and ended up agreeing on a very confident -3.

Based on the research above, we’ve produced a ‘base case’ for the decision aid. It is weighted heavily towards the beliefs of the meat eater in the collaboration since the question revolves around what a ‘typical’ person might think and meat eaters are more ‘typical’ than vegetarians. We would certainly encourage you to tinker with the worksheet yourself though, as some decisions are very personal. You can download it here.

From the model we get that the total impact of meat eating per typical western consumer is roughly -$9,500 – that is, the ‘society of conscious beings’ would be better off by around $9,500 per year if any individual human meat eater switched to eating plants instead. To put this in other units, it would be about as good for 5 people to go vegetarian for a year as it would be for medicine to extend one person’s life by one year.

Each value in the table below represents the annual impact of a decision to eat meat versus eating an exclusively vegetarian diet. The right-most attempts to express everything in the same units ($) based on a willingness to pay for a year of perfect human life of $50,000 and a year of YOUR OWN life of $100,000 (to reflect the fact that people generally care about their own welfare more, but if you are a perfect utilitarian feel free to set these both to $50k!). Per the model we find that even though cows, sheep and (very weakly) pigs prefer farming to non-existence, the number of chickens eaten and the conditions they are farmed in dominates the ethical considerations. In terms of other harms, the impacts on your health and the environment are moderate, and the financial impact of switching to a vegetarian diet is small but negative – that is, the typical meat consumer will in the long run prefer to eat meat than spend the savings from vegetarianism elsewhere.

Human
life-year equivalents

Expressed
in $ equivalent

Annual
impact on other conscious creatures

Cows

0.009

$373

Pigs

0.007

$317

Sheep

0.002

$65

Cage
chicken

0.000

$0

Shed
chicken

-0.138

-$5,913

Fish

0.000

$0

Environ.

-0.027

-$1,166

Impact
on you

Health

-0.039

-$3,336

Finance

 

$162

TOTAL
LIFE YEARS

-0.186

-$9,498

Overall, the impact of eating meat like a typical person is likely to be substantially negative. Eating no chicken limits the impact on the animals themselves, but the harm to your own health and the environment outweighs the moral good you do by causing the creation of animals who are happy at the margin.

7.1. Sensitivity

The decision aid allows you to specify uncertainty over any of your estimates, which we have done anywhere we are still uncertain about the value of a parameter. This analysis is displayed in the graph below. For any plausible distribution of inputs meat eating is harmful to you personally, primarily for health reasons; and meat eating generally causes harm to other conscious creatures because of the impact on environment and the high suffering of chickens. However, there is significant uncertainty about this value; in a small number of cases eating meat actually produces benefit to society by creating more lives animals would prefer to live on net.

Another way of exploring model uncertainty is scenario analysis. We’ve calculated a number of scenarios that cover likely areas of disagreement.

In order of least to most harmful to other conscious being, the scenarios are:
• No factory farming narrowly results in outcomes which favor eating meat, since every animal would prefer to be alive than not. The effect is not greater because there are still environmental and health implications to eating meat.
• Chickens not conscious – The base case assigns a 75% chance that chickens are conscious, and this is a big assumption to which the model is highly sensitive. Assuming chickens are not conscious results in outcomes which narrowly favor being a vegetarian, since the moral importance of creating worthwhile cow, pig and sheep lives is offset by the other harms of meat eating.
• Base case – As described in the document
• No meat causes depression – In this scenario not eating meat causes you a significant but not life-threatening illness – modelled by having minor depression (0.62 utility). This scenario is very interesting because it predicts that eating meat would be good FOR YOU, but would harm others, and therefore whether you should eat meat or not depends on the valuation you place on your own happiness versus the happiness of others (remember the model already values your QALYs higher than anyone else’s based on the assumption you are not a perfect utility maximizer AND you are compensated for the unhappiness meat causes).
• Environmental worst case – the resources used to create meat are the upper end of the plausible range (10%) and all of this resource will create new people. The more convincing you find the environmental argument the more likely you should be vegetarian
• No habituation – In this scenario no creatures (including humans) habituate at all, meaning they are exposed to the full ‘badness’ of the farming conditions. The less you buy the habituation literature, the more likely you should be a vegetarian; this is a very strong result
• No speciesism – In this scenario the value of conscious experiences for animals is weighted just as much as conscious experience for a human. This is the scenario that results in the strongest argument against eating meat, and could perhaps be the intuition driving the sometimes acrimonious state of discussion between meat eaters and vegetarians.
• Not shown in this analysis is an ‘Unfettered Vegetarian’ analysis, where the vegetarian collaborator is able to enter their own assumptions into the model without any check from the meat-eating collaboration partner. This is because what the vegetarian considers highly plausible assumptions (chickens are conscious, a much greater weight is placed on animal suffering/experience and much less habituation occurs) results in values that fall off the end of the graph – around $250,000 worth of harms to others per year.

Our key takeaway is that even under the most extreme scenarios we could think of meat eating is still very likely to be a net harm to both you and wider society. Also note that even in scenarios where you are not doing harm to both yourself and society, you are certainly hurting one of them quite a lot.

7.2. Impact on Collaborator Lifestyle

The meat eating collaborator was impressed by the environmental impact of beef and moral cost of factory farmed chicken. For the moment he has significantly reduced consumption of both, offsetting in part with salmon because fish have less environmental impact and are most likely not conscious.

The vegetarian was surprised how marginal the case for vegetarianism was when a ‘typical’ perspective was considered. Part of this is because he is still pretty skeptical that animals would actually habituate to the conditions we farm them in given the habituation literature doesn’t really cover conditions as cruel as factory farming. Another part might be that this collaboration has not focused on one-off traumatic events – especially slaughter – which probably don’t affect lifetime utility much but might be regarded as so self-evidently ‘evil’ that the way we have thought about the problem as a balance of good versus harm is incorrect. Having said that, although the actual harms of meat eating are less than he expected, the certainty of those harms occurring under any plausible distribution of beliefs (the fact that even a ‘typical’ person would probably regard meat eating as harmful considering everything) will probably make the vegetarian more militant about his vegetarianism. Sorry!

That being said, both collaborators agree that there is no substitute for evaluating the evidence for yourself. We can only hope that you find our analysis a useful reference.

This entry was posted in Uncategorized and tagged . Bookmark the permalink.

416 Responses to [ACC] Is Eating Meat A Net Harm?

  1. RLM says:

    Try noticing what happens when you read the sentence “The dustmen said they would refuse to collect the refuse without a raise.” Notice how the word ‘the’ appeared twice? And how you read it both times despite Scott’s best efforts at conditioning you otherwise? Very good.

    … I guess we’re missing an extra “the” in that sentence? (Or I’m VERY bad at noticing two of them)

    • Scott Alexander says:

      I am also confused about this.

      • Ashley Yakeley says:

        I can verify that the word “the” appears exactly twice in the sentence “The dustmen said they would refuse to collect the refuse without a raise.”.

        • Said Achmiz says:

          Sure, but this banal fact does not seem to serve the authors’ point in any way I can discern.

          • Taleuntum says:

            Maybe this way: You could have skipped a “the” in the sentence (as you routinely do, because of Scott), but you did not and this decision to not skip was not conscious on your part, your brain did this all by itself. (this supports the previous topic sentence that there are things the brain does which do not give rise to conscious experience)

      • Tom Chivers says:

        I had assumed that the editor, ie you, had edited it out and immediately ruined the point, in the way that editors do. But since it’s a point you make often yourself, that would seem unlikely.

        So my new hypothesis is that they forgot to put the extra “the” in there. I reckon you can reinstate it. Or ask them, maybe.

      • Robert Jones says:

        It’s a joke based on the fact that in this blog “Try noticing what happens when you read the sentence…” is often followed by a sentence containing a repeated “the”. The authors are humorously implying that SSC readers may have been conditioned not to read the second “the” even though they appear here non-consecutively, while also referring back to another example of the brain tidying things up before they reach the level of consciousness.

        • johnhancock says:

          Thank you for explaining this. I vaguely understood an attempt at a reference to the whole “the the” brain-illusion meme there, but couldn’t for the life of me make any sense out of it beyond that.

      • j1000000 says:

        I believe they’re pointing out the two different appearances of “the” at different places as being a simple thing for the brain to interpret, while the two appearances of “refuse” are in theory harder for the brain to parse but it still does it automatically. And the joke is presumably as if this case of two the’s several words apart is in any way like what Scott does.

      • thevoiceofthevoid says:

        Additional typos:

        Slaughter chickens in the West are raised instead in a large ‘broiler’ shed covered with liter.

        In a study by Kestin, between 2% and 30% of broiler chickens, depending on the breed, had a gait score between 3 and 5 on a 5 point scale (1 no issues, 3 obvious gate defect, 5 unable to move at all).

    • ACC2019 Omnivore says:

      It’s a joke. Written as intended.

  2. HarmlessFrog says:

    Disclosure: I don’t eat plants for health reasons.

    Whether or not a plant based diet is viable from a nutritional perspective depends mostly on whether you have the economic means to consume a wide variety of food sources, and may be riskier for small children or those whose ancestry is from regions where meat-eating was prevalent.

    This matches to “everywhere”. What, indeed, were people eating before agriculture? What were they eating before dairying?

    4.2. Long Term Health Outcomes

    This whole section is “garbage in, garbage out”. All of these associations are weak or non-existent. Take a look at a EPA publication on interpreting epidemiology. It’s extremely unlikely that meat is the causal factor.

    Also take a look at the review of evidence against meat. It’s ridiculously weak, too weak to tell people to even limit, never mind give up, meat consumption.

    In total, meat, aquaculture, eggs, and dairy use ~83% of the world’s farmland and contribute 56-58% of food’s different emissions, despite providing only 37% of our protein and 18% of our calories.

    How much of that farmland is actually farmland that you could grow some useful crops on? There are far more places where you can effectively raise animals, but cannot effectively farm plants, than places where you can do both. I highly recommend dr Ballerstedt’s work on the subject of ruminant agronomy.

    Of course, humans don’t eat exactly the same food animals eat, and vegetarians are for some reason unwilling to just drink 2000 calories of canola oil every day.

    One reason is that it’s ultra-processed industrial junk. I would sooner burn it for heat than give it to an animal to eat, much less a human.

    Overall, the case for reduced meat consumption is strong. Vegetarianism is cheaper, better for your health (if you can afford a diverse diet and are not an infant), and is less impactful for the environment. It also has a significant moral cost in terms of animal suffering.

    No, on every count except cheapness. It is cheaper to eat a nutritionally-deficient, industrially-extruded diet. It is hardly good. Giving up meat is something to do if you think a high prevalence of stunting is an acceptable price to pay.

    I would personally invite both authors to leave the horrid field of nutrition for a moment, and take a look at paleoanthropology. Humans, ever since they have been humans, have eaten meat. This is the probable cause as to why we have even developed such an extremely energy-hungry brain, while reducing the size and power of our digestive system. It is ludicrous to suggest that humans are somehow better adapted to not eating meat, based on the scant evolutionary time afforded them since the advent of agriculture, compared to the multi-million-year history of meat consumption.

    The problem with public health, as I see it, is that we’re eating far too little meat and animal fat… and far too much processed garbage, especially industrially-produced seed oils.

    • HarmlessFrog says:

      Also, the protein requirements for adults have been underestimated, so it’s not just infants who fail to get their optimal amino acids in:
      https://www.ncbi.nlm.nih.gov/pubmed/19841581
      https://www.ncbi.nlm.nih.gov/pubmed/26962173
      https://www.ncbi.nlm.nih.gov/pubmed/25320185

    • Unirt says:

      I don’t disagree with this, but for a counterargument remember that diets richer in amino acids are associated with lower life expectancy in most animals from worms to insects to rats.

      There are also the interesting findings that some species switch to more protein-rich diets in adverse environments, e.g. when they perceive high predation risk, leading to faster maturation. Nitrogen compounds can be slightly poisonous, even if also necessary. It looks like the added poteins maybe enable one to live faster and die younger, which can be a wise thing to do with many predators around. This is a wild hypothesis, though.

      • HarmlessFrog says:

        I don’t disagree with this, but for a counterargument remember that diets richer in amino acids are associated with lower life expectancy in most animals from worms to insects to rats.

        Association is not causation, non-human animals aren’t humans.

        • Unirt says:

          Association is not causation, non-human animals aren’t humans.

          Sure, but the relationship is so prevalent that it may easily have something to do with quite fundamental chemistry inside animals, in which case it wouldn’t be easy to escape it. I would at least consider this as a point against meat eating, if not a particularly strong one.

          • HarmlessFrog says:

            I’m personally not convinced overmuch by the mToR-phobic, anti-protein argument, on grounds that humans used to eat a lot more protein, yet did not have significant chronic diseases until the advent of civilization. Furthermore, evolutionary evidence suggests we are especially well adapted to living on our protein ceiling as established by MRUS.

          • Unirt says:

            Yes humans and cats have certainly evolved to eat a lot of meat, but it’s not clear in humans (and not even in cats) what is the optimal ratio of proteins vs other stuff in diet. I don’t have a strong opinion on it.

            did not have significant chronic diseases until the advent of civilization

            This is probably factually wrong. There certainly weren’t many obesity epidemics, but significant cronic diseases toward the later halves of their lives are tragically common among mammals, while not so much among birds.

          • HarmlessFrog says:

            did not have significant chronic diseases until the advent of civilization

            This is probably factually wrong. There certainly weren’t many obesity epidemics, but significant cronic diseases toward the later halves of their lives are tragically common among mammals, while not so much among birds.

            You have sources? I definitely would like to see evidence of similar rates of cancer, diabetes and cardiovascular illness among hunter-gatherers as among civilized people. Even data from primitive agriculturalists would be interesting; I have directly opposing evidence. According a contemporary hypothesis that I personally find convincing, obesity is exactly the problem, whether visible or hidden (as in lipodystrophy). Exceeding the maximum storage capacity of your adipose tissue causes insulin resistance, and insulin resistance gives you chronic disease, via sustained hormonal imbalance and energy toxicity.

            https://www.ncbi.nlm.nih.gov/pubmed/30511505
            https://www.ncbi.nlm.nih.gov/pubmed/10535381
            https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)30752-3/fulltext?elsca1=tlprhttp://www.thelancet.com/journals/lancet/article/PIIS0140-6736(17)30752-3/fulltext?elsca1=tlpr
            https://www.ncbi.nlm.nih.gov/pubmed/25515001
            https://academic.oup.com/jcem/article/86/8/3574/2848584
            https://academic.oup.com/jcem/article/83/8/2773/2660508
            https://www.ncbi.nlm.nih.gov/pubmed/26401581

          • Unirt says:

            I definitely would like to see evidence of similar rates of cancer, diabetes and cardiovascular illness among hunter-gatherers as among civilized people.

            Sorry I was too vague; I’m not trying to argue with this. Of course obesity is a major player that strongly increases the risks. I’m just saying that cancer, arthritis, dementia etc are all present in wild animal populations (but the affected animals and probably also pre-modern-medicine humans die fairly quickly after developing them). It’s not strictly correct to say that chronic diseases were insignificant before civilization.

          • HarmlessFrog says:

            Sorry I was too vague; I’m not trying to argue with this. Of course obesity is a major player that strongly increases the risks. I’m just saying that cancer, arthritis, dementia etc are all present in wild animal populations (but the affected animals and probably also pre-modern-medicine humans die fairly quickly after developing them). It’s not strictly correct to say that chronic diseases were insignificant before civilization.

            Animals vary greatly in rates of cancer. For mice cancer is a given, not so for elephants. What is noteworthy is that there is a human population that suffers from very little cancer (95% reduction), despite having similarly high other chronic diseases – Laron syndrome dwarfs (one of the links above is about them, they have normal insulin, but very little IGF1). Explorers such as Stefansson and Price noted that colonial physicians, who had Caucasians and native populations living in close proximity, reported extraordinarily low rates of cancer in the natives (while they gave white man’s food a wide berth), while the colonials had plenty. These data points suggest to me that healthy humans have a very low baseline cancer incidence.

            This is backed up by one of those SSPG studies (also linked above), where the only cancers occurred in the second and third tertile by insulin resistance. The same can be seen for other chronic diseases – the healthiest tertile by SSPG does not develop them at detectable rates at the sample sizes. If chronic diseases were reduced to a single digit occurrence, that, in my mind, would make them insignificant compared to what we have today, with two-thirds of the civilized population being insulin resistant and heading for an early death.

          • Unirt says:

            Thank you, these links are interesting.

            I wonder what hunter-gatherers die of, if it’s not age-related chronic diseases. This leaves only infections, accidents and violence? Or perhaps they still die of age-related diseases, but at older age?

          • HarmlessFrog says:

            You’re welcome.

            Infections, accidents and violence are the bulk of it, AFAIK. They still grow weak and decrepit in old age, and die of that; entropy still wins.

    • NostalgiaForInfinity says:

      What is the evidence for what proportion of prehistoric human diets were meat? My understanding is that contemporary hunter gatherers eat surprisingly little meat (except for some small populations e.g. the Inuit). But obviously that isn’t necessarily the case before the development of agriculture.

      • HarmlessFrog says:

        What is the evidence for what proportion of prehistoric human diets were meat?

        To my knowledge, there is no quantification possible on the basis of existing evidence; there’s plenty of evidence that we did eat meat everywhere for a long time, but not a lot of direct evidence of how much. Isotope studies would place early humans at about the same trophic levels as known top-level carnivores. The anatomy of our digestive tract suggests a diet similar to a wolf or vulture. The size of our brain practically demands a diet of high quality, doubly so because of our rather unimpressive maximum rate of urea synthesis. The rather conspicuous holocene extinctions of large (and therefore fatty) animals is also a strong clue.

        My best guess would be somewhere upwards of 70% of calories by the time of Homo erectus.

        https://www.nature.com/scitable/knowledge/library/evidence-for-meat-eating-by-early-humans-103874273/

        My understanding is that contemporary hunter gatherers eat surprisingly little meat (except for some small populations e.g. the Inuit).

        Recent HGs eat/ate far more meat than we do, on average. How much they eat depends on the environment. The colder it is, the less edible plants there tend to be. Given the long spells of colder climate in the last few million years, I expect this would have extended to much of Africa as well.

        https://www.ncbi.nlm.nih.gov/pubmed/10702160

      • Nornagest says:

        I can’t be bothered to dig up a cite, but I dug into this a while ago and found that contemporary forager cultures usually have diets ranging from about 30-70% animal (not quite identical to “meat” — foragers rarely eat dairy for obvious reasons, but eggs aren’t uncommon) if you drop the Inuit and related cultures (for whom it’s almost 100%) as an obvious outlier. There are huge variations in what types of animal, though, between cultures and seasons — for one culture that might be mostly shellfish year round, another might spear huge numbers of salmon in season and rely on grubs and bird eggs in the off-season, a third might routinely snare small woodland critters and occasionally bring home a stag.

        There are always some caveats when you’re talking about contemporary foragers, who’ve almost all been pushed into marginal land and frequently have cultures affected by interchange with neighboring sedentary societies, but the overall pattern’s pretty consistent.

        • HarmlessFrog says:

          That’s Cordain’s study I linked in my top-level post (second to last link). And that one probably underestimates the amount of animal dependency, IIRC, because it assumes that everything gathered is plants, and that’s not true.

    • onyomi says:

      What, indeed, were people eating before agriculture? What were they eating before dairying?

      Fruit, nuts, seeds, roots, greens, honey… (in addition to some meat when the hunt/fish was successful)?

      • onyomi says:

        Or whatever they could find before someone else got it.

        • HarmlessFrog says:

          I believe that’s called the “optimal foraging strategy” and explains well why humans do so poorly in the modern environment.

      • HarmlessFrog says:

        I don’t mean to suggest that they didn’t eat plant foods. I’m suggesting that they relied substantially, even primarily on meat, particularly in seasonal and northern climates. I don’t think you will find any vegetarian or vegan populations of the genus Homo prior to agriculture, but I won’t discourage you from looking.

    • Gabe says:

      With all due respect to the authors, when I think of “adversarial collaboration,” I think of someone like Harmless Frog here teaming up with a hardcore vegan to get to the bottom of things.

      • caryatis says:

        I agree with this. This wasn’t adversarial enough. Also, the topic was too big. You could do a lengthy collaboration on just human health impacts of meat, or just environmental impacts of meat. The ethical questions are so highly dependent on whether/how much you value animal interests that i question whether they can be usefully addressed in this format.

      • HarmlessFrog says:

        Any hardcore vegans around? 🙂

      • JohnBuridan says:

        Scott explicitly asks that all the Adversarial Parts happen under the hood and the resulting collaboration is a united front. So the results will always look less adversarial.

        I think the better method of presentation would be a Scholastic question or Platonic Dialogue in which all the adversarialness in on the table.

    • thevoiceofthevoid says:

      How much of that farmland is actually farmland that you could grow some useful crops on? There are far more places where you can effectively raise animals, but cannot effectively farm plants, than places where you can do both.

      I think the figures include crops raised to feed animals–IIRC it takes about 10 calories of plant feed to produce 1 calorie of beef. So, instead of using the vast tracts of farmland to grow feed corn, we could hypothetically use them to grow e.g. corn for human consumption.

      • HarmlessFrog says:

        I think the figures include crops raised to feed animals–IIRC it takes about 10 calories of plant feed to produce 1 calorie of beef. So, instead of using the vast tracts of farmland to grow feed corn, we could hypothetically use them to grow e.g. corn for human consumption.

        When you grow human food (such as corn, I guess), you also produce a mess of extra material that is not edible by humans, but is edible by livestock. And this is exactly what they do in the industry. Do you think they would just throw away everything that’s not the cob? That would be completely wasteful and financially inefficient.

    • thevoiceofthevoid says:

      Of course, humans don’t eat exactly the same food animals eat, and vegetarians are for some reason unwilling to just drink 2000 calories of canola oil every day.

      One reason is that it’s ultra-processed industrial junk. I would sooner burn it for heat than give it to an animal to eat, much less a human.

      far too much processed garbage, especially industrially-produced seed oils.

      Cool video! The thing is, unless there’s a specific toxin present at a biologically relevant level, I think you’re committing a naturalistic fallacy here. Canola oil is made of the same kind of fatty-acid hydrocarbons as olive oil or beef tallow, albeit in different proportions. (Canola oil actually has less of the varieties [saturated fats] generally agreed to be bad for you.) The point of all the industrial machines is just to extract those compounds from the seeds, how exactly do you think the processing turns it into “garbage”?

      • HarmlessFrog says:

        Cool video! The thing is, unless there’s a specific toxin present at a biologically relevant level, I think you’re committing a naturalistic fallacy here. Canola oil is made of the same kind of fatty-acid hydrocarbons as olive oil or beef tallow, albeit in different proportions.

        You might as well claim it’s made of the same atoms, which would be true, but useless. The proportions of different fatty acids are important, very much so in this case.

        And you’re actually wrong – rapeseed oil and olive oil actually lack two polyunsaturated fatty acids humans use, DHA and EPA; the plant version of the omega-3 fatty acid is ALA, which is poorly converted into things we can actually use. Given their abundance in omega-6 fatty acids, that makes the ratio to omega-3 particularly bad. The omega-3 index is has a very strong association with all kinds of bad things. Yet, fish oil supplementation interventions have very modest, or non-existent effects. To me this looks like the omega-3 index is actually an index of how much vegetable oil you are NOT eating.

        https://www.ncbi.nlm.nih.gov/pubmed/9637947
        https://www.ncbi.nlm.nih.gov/pubmed/19852881
        https://www.ahajournals.org/doi/10.1161/JAHA.119.013543

        (Canola oil actually has less of the varieties [saturated fats] generally agreed to be bad for you.)

        Saturated fats are not bad. If you think so, you have not read the literature. To be fair, neither have the people who make the recommendations to avoid saturated fat, apparently. In fact, my current hypothesis as to the crux of the modern health problems is precisely the replacement of natural animal fats with industrially produced vegetable oils. The condemnation of saturated fats was never based on an honest review of the literature. We have data going back to the 1960s showing that replacing saturated fat with polyunsaturated seed oils is not beneficial and likely harmful. There’s good reason to believe that replacement of saturated fat with unsaturated fat, particularly linoleic acid, is one of the primary causes of the obesity epidemic.

        https://www.youtube.com/watch?v=NUY_SDhxf4k
        https://www.bmj.com/content/353/bmj.i1246
        https://www.bmj.com/content/346/bmj.e8707
        https://www.youtube.com/watch?v=pIRurLnQ8oo

        The point of all the industrial machines is just to extract those compounds from the seeds, how exactly do you think the processing turns it into “garbage”?

        Aside from problems such as using neurotoxic solvents like hexane for extraction, the refinement itself is a problem. It takes something humans would not normally be able to ingest and absorb in nearly as large quantities, and makes it easily ingestable and absorbable. Humans have no likely adaptation to eating that much linoleic acid. Particularly when such oils are used for cooking and when they are exposed to long term storage, given the instability of PUFA.

        https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5093368/
        http://wholehealthsource.blogspot.com/2008/09/pracical-approach-to-omega-fats.html

    • Gumpalonia says:

      I also concur with the complaints about this appearing like a collaboration between two ideological vegans, although one of them might be eating meat (but obviously feeling terrible about it). Even the whole frame is biased: like we have a choice between harming animals and not to harm them. But do we even have the choice to grow any food without killing animals? Doesn’t growing monocrops cause the deaths of many sentient beings like mice, lizards and birds? So wouldn’t an more honest discussion be about what animals to kill and whether to do it directly or indirectly?

      Also the stories about when animals feel their lives are worth living and when not can be created in more ways than were presented. One could think that grass fed cows are actually living in bliss like conditions far superior to their wild cousins. Just think about the conditions they face in the wild. Constant danger and predation causing stress, seeing your tribe members being eaten alive and (if having consciousness), knowing that will be your end as well. Now to take an analogy to humans, imagine living as a human, being hunted by aliens:

      (SFW) https://en.wikipedia.org/wiki/Alien_(creature_in_Alien_franchise) , and going through what a cow goes through from wild predators. Roaming with your tribe and knowing there is someone trying to eat you whenever you turn your back. You could never defeat them, just outrun, for a moment. They would always be there lurking after your flesh, seeing you as nothing but calories. And contrast that with a life being herded by, let’s say, by predators (SFW): https://www.google.com/search?client=firefox-b-e&biw=1409&bih=930&tbm=isch&sa=1&ei=aWLzXfSFPO7yqwHE9Z_wCA&q=predator+&oq=predator+&gs_l=img.3..0i67l10.401622.403997..404915…1.0..0.71.909.14……0….1..gws-wiz-img…..0..0.FWJk_WnZmWA&ved=0ahUKEwi03raDsLLmAhVu-SoKHcT6B44Q4dUDCAY&uact=5#imgrc=H4LcPWr5wgxqSM: You would live in an environment naturally suited for you, all your needs would be catered for. You would maybe see a fence somewhere, but you wouldn’t know if it was there to keep you from escaping or other creatures from hurting you. You would see people taken away when they reach middle ages, by predators, but you wouldn’t see them harm anyone. Sometimes they would even be affectionate towards you. You might think they are your servants, and they would make your existence stress- and carefree.

      The question is which life you would choose? (If you think this analogy is bad, that´s the point, analogies are bad and presuasive only when certain priors are shared).

  3. Lipman says:

    However, in practice animals are often not insensible to pain when they are skinned or carved up, […] because of […] religious beliefs around the way meat should be prepared (link, further figures)

    The links say nothing of the sort. If anything, the opposite is stated, pointing to claims “that stunning itself can cause great suffering”.

    In fact skinning or carving an animal that isn’t dead would render it forbidden for consumption by Jews (and, I suppose, by Muslims, but I’d have to google that as much as the authors). It’s an absurd thought. The whole idea of Jewish slaughter is that the animal loses consciousness instantaneously, and this has been shown to be the case in studies.

    (For what it’s worth, I’m a vegetarian.)

  4. Said Achmiz says:

    This AC, like the circumcision one, was disappointing (for all that it was longer and had all those links and numbers). As with the previous one, it seemed like one side was simply not well-represented. The impression I got was that one person was an avowed vegetarian and the other person was just some vaguely-“rationalist” sort who ate meat and thought that eating meat was okay but didn’t really have any strong opinions on the matter. As someone who (a) does eat meat, and (b) has thought for longer than five minutes about the moral issues involved, the bulk of this write-up strikes me as both faintly ridiculous and, more importantly, irrelevant. (As for the health and environmental issues—other commenters have, thankfully, already noted the highly questionable nature of the given claims.)

    I should have liked to see a collaboration on this subject between two people who actually started with strong, principled, and genuinely opposing views. Alas, that is not what we appear to have gotten.

    • HarmlessFrog says:

      I should have liked to see a collaboration on this subject between two people who actually started with strong, principled, and genuinely opposing views. Alas, that is not what we appear to have gotten.

      I would pay to see a collaboration between dr Michael Greger and dr Shawn Baker, the risk of matter-antimatter annihilation be damned.

    • jasmith79 says:

      I don’t disagree with you, and I wouldn’t tell you not to say it.

      But:

      …I feel compelled to point out that the most likely alternative to this collaboration is not a better collaboration, but the toxic slew of slogan-chanting zombies “arguing” on social media (i.e. no collaboration).

      I think this collaboration has value, lots of value, and I’m very grateful to the authors for all of the work they put in and to Scott for sponsoring it’s creation.

      • Randy M says:

        Agreed. I don’t take it as definitive but surely valuable.

      • sclmlw says:

        Given that this is essentially a one-sided report, I don’t agree that it’s better than nothing since it’s strongly prejudicial while carrying the name of an ‘adversarial collaboration’ and it doesn’t really feel like there was an adversary on the other side.

        This was written in a strongly biased tone. Some of their framing is far afield, and they try to make a quantitative analysis out of things that are obviously qualitatively different without critically assessing whether that’s a worthwhile project. It felt like the net-harm side got a free pass to make their case unopposed.

        I get the impression that if strong opposition had been present we would have ended up with no submission overall, since most of the conclusions are challenged in the comments here.

        If you want to hear a strongly biased argument on this topic (or any other) that is not difficult to find. The purpose of the ACC was to outline where the arguments overlap in the Venn diagram so we can better define the debate. I don’t feel like that was done at all, so this was not a successful collaboration. I’d have preferred a collaboration that failed to produce a final product to this one, since at least then I’d know the Venn overlap was near-nonexistent. This doesn’t tell me anything I didn’t already know.

      • jasmith79 says:

        this is essentially a one-sided report, I don’t agree that it’s better than nothing

        I don’t necessarily disagree with that either, the patina of neutrality is handy flinging feces for the side that “won”. But that being said, again, this piece is, and this is germane to the point that I was trying to make above, completely open and transparent about it. The motivations and dispositions of the authors are explicated! It not only meets the bar of having a transparent methodology but exceeds it, which is why I’m glad this exists.

        It makes some important (although not enough) points about nutrition density, the moral value of factory farmed cows, etc.

        I don’t want to forestall necessary and valuable criticism, but let us do so in the proper spirit: this is still better in many ways than a normal meta-study and is the Virgo Supercluster of respectability v. Tumblr.

    • DragonMilk says:

      Unfortunately, I think many ACs are the product of someone with a strong opinion and a, “meh, whatever” dude who tags along.

      It takes much more time for genuine adversaries to come together and write a unified report.

      I think it may be better for Scott to dictate the structure to be as follows:

      1. Introduction of authors and viewpoints
      2. Author A presents the case for X
      3. Author B presents the case against X
      4. Authors A and B discuss attempts to reconcile differences and and find common ground
      5. Conclude with areas to where they agreed to disagree and next steps for potential future collaboration/discussion.

      Otherwise, a lot of these seem to focus a lot more on a mix of 2 and 4.

    • dionisos says:

      It doesn’t seem to me that one side was not well-represented.
      It seems there were strong disagreement on a lot of things, and they ended up representing a lot of possible positions and quantitative values on what they found controversial.

      Not shown in this analysis is an ‘Unfettered Vegetarian’ analysis, where the vegetarian collaborator is able to enter their own assumptions into the model without any check from the meat-eating collaboration partner. This is because what the vegetarian considers highly plausible assumptions (chickens are conscious, a much greater weight is placed on animal suffering/experience and much less habituation occurs) results in values that fall off the end of the graph – around $250,000 worth of harms to others per year.

      I believe this say something.

      Maybe your position weren’t well represented, but this is a entirely different thing.

      • Said Achmiz says:

        Yes, your quote does say something. And that something is exactly what I said: that the vegetarian position was massively favored over the non-vegetarian one. (The latter being, indeed, “my” position… and also one of the two positions that was supposed to be represented in this AC—but wasn’t.)

        • dionisos says:

          I don’t get it, there is a factor of almost 40 between the vegetarian position and the one they put in their article.

          • Aapje says:

            What is the factor between the the assumptions of the model and a plausible meat-eater position, like: the suffering is far less than claimed & less than in nature?

            We don’t know, because only the vegetarian apparently pushed to get the model far more in their direction, but not vice versa. This suggests that one side was far more influential.

          • dionisos says:

            We don’t know,

            The authors can’t know more than we.

            It is a collaboration between one meat eater and one vegetarian, and I believe the numbers wasn’t closer to the vegetarian position than the other one.
            And the best they can do to deal with other values was to let people change the numbers to get the result from their own position, what they did.
            The first ACC was on-sided in my point of view, but this one is not, because of this. (they tried to encompass a great range of considerations and positions with their framework)

            It feel like you are asking for the meat-eater to not only represent his own position, but to represent a sort of global meat-eaters position.
            But it is hard enough to meet some common ground without adding this difficulty.

            Or you think the meat-eater didn’t hold a classic meat-eater position in the first place.
            Maybe, maybe not, really hard to know.

            It could be really interesting to do a redo of it with a more “hardcore” meat-eater.

            less than in nature

            A lot of vegans I know and a lot meat-eaters seems to hold the position that factory farming is worst than life in nature and that other kinds of farmings are better.
            But I am very probably in a bubble, I don’t know what is the general position on this.
            (anyways in my point of view this consideration isn’t really pro-meat eating or pro-vegetarianism).

          • Aapje says:

            No, what I ask is for the meat-eater to be as critical of the vegetarian claims as seems to be the case vice versa.

  5. gkai says:

    Interresting take, the diet and environmental effects are often debated and the analyse follow a classic path…Not to say conclusions are clear, but it looks like a more or less objective conclusions could be reached, at least in principle. On the animal suffering and especially link with animal consciousness, I am much more doubtful. It seems highly speculative, even saying that you need consciousness to suffer seems strange. The sorting of animals is already not easy from an intelligence point of view, but if you try to split between intelligence and consciousness I fear it become largelly subjective…In particular, human bias will play a large role in pushing up animals who trigger sympathy (looking like infants, soft looking, species whose body langage is easier to interpret, or whose instints are closer to human ones)… People will certainly rank koala suffering much much higher than crow, or even more, octopus, but should they? Maybe the only animal value is arbitrary human perceived value anyway, and the rest is rationalisation
    Also, for farm animals, I think habituation can not be considered if the selection is not taken into account. The domesticated version hugely differ in behavior from the wild version, and I think this play a much bigger role in term of comfort felt in various condition than individual habituation. You could select for happiness in farm condition, and to a certain extent, it was indeed selected. Including selection in animal well being discussion makes it even more difficult, but imho it is a must, it’s really the elephant in the room…

    • waterwyrm says:

      The economist Mike Munger has a post from 2014 that begins:

      Unicorns, of course, are fabulous horse-like creatures with a large spiraling horn on their forehead. They eat rainbows, but can go without eating for years if necessary. They can carry enormous amounts of cargo without tiring. And their flatulence smells like pure, fresh strawberries, which makes riding behind them in a wagon a pleasure.

      He goes on to argue that we end up arguing not about things, but our idealization of the things.

      He proceeds to quote Edmund Burke:

      It was observed, that Men had ungovernable Passions, which made it necessary to guard against the Violence they might offer to each other. They appointed Governors over them for this Reason; but a worse and more perplexing Difficulty arises, how to be defended against the Governors?

      Burke is already off the rails: the new problem is indeed more perplexing, but it’s mainly worse only if you are ideologically confounded by attempting any solution at all. I guess he prefers a system where the governed have no set consensus about who among them functions as the arbiter of last resort. Perhaps we can at last solve this vexing problem with the digital mechanism of the blockchain distributed ledger. Too soon to say. But finally, at last, a glimmer of some other viable path forward.

      Governance does indeed get us all hot and bothered, but at least it’s not always a conceptual unicorn. It is possible to have a conversation about the real world in this domain, however difficult it might be to stay on the rails of the actual or the actualizable.

      Not so with consciousness or free will. God invented philosophy so that conceptual unicorns could be discussed in polite company. Three biggest unicorns in the known universe: what is consciousness?, in what form does free will exist? (if at all), and why something rather than nothing? (should the concept of “nothing” someday prove to be well-defined).

      At computers entered the public consciousness back in the 1970s—the Apollo program was omnipresent—the first item of hard lexicon—the very first digestive biscuit for the digitally unwashed—was GIGO: garbage in, garbage out. You see, computers are error free, this being found money of a fully transcendental order:

      On two occasions I have been asked, “Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?” … I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.

      — Charles Babbage, Passages from the Life of a Philosopher

      Clearly we remain in dire need of updating this old saw as follows: Unicorns in, unicorns out.

      Two smart, industrious souls take their pink, perfume-tailed unicorns, add them together, and divide by two (black hat required, may be absent as illustrated). Presto bingo, baron of hoof with a strawberry remoulade.

      Apparently, this simple lesson passes the smell test, but does not pass the spit test: discussing UIUO soon spreads a fine mist over the entire affair. In all matters of haute couture, one first consults the Japanese. So I propose instead to use the Japanese word “kirin” in place of “unicorn”, rendering our urgent philosophical maxim as KIKO instead, which one can now discuss in elevated language, while consuming artful morsels of nearly raw eel in a shatteringly crisp tempura batter.

      Just a few weeks ago, aldaily.com syndicated a ludicrous article from the philosopher Jim Holt. This starts well—John Stuart Mill suggesting “it is better to be a human being dissatisfied than a pig satisfied”—and ends well—with a cryptic reference to Nabokov’s sophisticated autobiography Speak, Memory—but leaves me shaking my head over nearly everything in between. If I didn’t know Holt already from another work, I would have been sure the Nabokov reference was intended to stick a pin into the entire fat balloon. But, unfortunately, I do know Holt from another work: his book, Why Does the World Exist? (2012) about which I noted at the time: “he seems to evade the central question, in chapter after chapter, interview after interview”. The central problem being that he desperately seeks evidence for an actual unicorn.

      I’ll pass off finding my own words concerning this “existential detective novel” (as promoted) to Manny over at Goodreads: Jessica Q. Rabbit, singer, model, movie star and high-flying academic, talks candidly to The Toon Town Times about Why Does the World Exist? who ultimately signs off “I feel I’ve been such a bitch talking about Jim this way, I mean he’s a nice guy and all and I’m sorry about his dog and his mother, but you know, he just totally pushes all my buttons.”

      Returning to GIGO from the Apollo era, there we find Thomas Nagel’s famous paper from 1974 “What is it like to be a bat?” in which he argued that an organism had conscious states “if and only if there is something that it is like to be that organism—something it is like for the organism.”

      Suppose that it is, indeed, “like something” to be a human being, “like something” to be a bat, and “like something” to be a pig. This gets us precisely nowhere—unless you also place faith in conceptual unicorns: that all these “like somethings” have some essence in common about which we can confidently proclaim, from our little corner of the universe, in which we are only and exclusively “like something” of subtype human being. If it happens that the human subtype does not fall far from the prototype tree, perhaps we can infer something about the pig, as well. Or even the chicken. The vogue these days is to feel skin-shame over our sad and brutal legacy of cultural imperialism in the human domain. Meanwhile, in the sphere of projecting our own “like something”ness onto our many other arkmates, our wokeness has yet to progress much beyond the multitudinous voice-appropriation of Dr Doolittle.

      It’s completely fair game to look at our own face in the mirror, to inquire about the nature of our deep predilection to seek our own reflection everywhere. Jennie Erin Smith covers this ground wonderfully in Cows with character (March 2018) on the TLS (subscription required). Dubious it remains, however, whether we are seeing the animals in their deepest “like something”ness, or merely grazing upon a coterie of animal-shaped strawberry-flavoured unicorn cookies.

      The miracle of this exercise, it seems to me, is how the arithmetic mean of two unicorn cookies somehow seems instantly more credible than either unicorn cookie alone. “One does not simply walk into Mordor. Its black gates are guarded by more than tiny semantic quibbles. There is mystery there that does not sleep, and the Great Quarrel is ever recycled. It is a barren Zoroastrian wasteland, riddled with earth, water, air, and fire. The very water you swim in or air you breath is a perfidious wraith. Not with 10,000 Swedish Chefs could you take the arithmetic mean of two unicorn cookies and come up with a better answer than either cookie alone. It is chocolate upon moose to the highest order.”

      Swedish Chef prepares chocolate moose

  6. Taleuntum says:

    A humorous example of a glitch in bird behavior can be found in YouTube videos where the ‘imprinting mechanism’ of ducklings has confused them into thinking a dog is their mother.

    This is also a funny example of a glitch in chickens.

    Typo:

    There is good reason to believe all common land-based food mammals (cows, pigs, sheep, goats) are highly consciousness.

  7. Michael Watts says:

    The central question is whether factory farmed animal lives are worth living; the realistic alternative to meat eating is not a better life but for those animals to not exist in the first place.

    Is this really more central than the question “Why would we, not being cows, care what cows want?”?

    • HarmlessFrog says:

      An excellent question. I would answer, that it matters to the extent to which caring what cows want yields better meat and dairy for our consumption.

      • dogiv says:

        Interesting that you feel strongly on two points in favor of eating meat, both that the health effects are positive, and that animal suffering does not deserve our moral consideration. These two axes are roughly orthogonal, so is having extreme positions on both a coincidence, or does your conclusion motivate your reasoning on one or the other? (this is meant as a question, not an accusation)

        I do find your arguments about health effects somewhat convincing though I suspect people who also care about animal suffering could get all the benefits of animal consumption from fish and other non-“conscious” creatures.

        • HarmlessFrog says:

          These two axes are roughly orthogonal, so is having extreme positions on both a coincidence, or does your conclusion motivate your reasoning on one or the other? (this is meant as a question, not an accusation)

          The moral position is the pre-existing one, which has not changed, for as long as I remember.

          I used to have no particular opinion (being ignorant) on the subject of the health effects.

    • melolontha says:

      Do you doubt that cows are capable of suffering, or do you just not care? If the latter, what kind of argument could conceivably convince you? I think if you’re just missing that moral foundation, it would probably be a waste of time arguing about it.

    • Unirt says:

      Do you perceive your question as different from or analogous to asking “why would we, not being children, care what children want?”?

      • gkai says:

        Not sure for the original poster, but for me they are very different:
        – I was a children, I never was a cow. Related to that, I know how children think and feel (I know how I felt, and other children are probably similar). For cows, not so much, I may try to guess but it’s not easy, soo I care more about children….

        Same reason for caring more about fellow humans than animals, and more about chimps than shrimps: empathy is closely linked to sameness, you need to imagine switching places….and this is easier the more alike you are to the victim…

        -Children are special (your own children at least), there biological altruism is at play and you don’t even need any philosophical consideration: you are (at least most people are) genetically programmed to care.

        • Unirt says:

          I heavily anthropomorphize cows of course but to me it doesn’t feel difficult to imagine being a cow – it’s quick and automatic to do so. I also think I’m not that drastically off the mark (though I am somewhat, I know). Cow children behave very similarly to human children in several ways, so there are probably relevant similarities in their perception of the world too.

          I suppose, whether the question of why should we care makes sense to you or not depends a lot on how readily you imagine being a cow?

          • gkai says:

            yes, and how emphatic you are in general.
            I have both low empathy, and have trouble understanding what it would be to be a cow (being inside a cow body is easy to imagine, but that’s very different – it would be a human-brained cow) so yeah, eating a steak does not trouble me much.

            Chicken is even worse, I can kill them myself if there is need, not pleasant but not something that troubled me much either. I doubt I could kill a cow myself and not think much about it…

            If you want to go the the far end of things, there is mosquitoes. There it is pleasant to kill them, after they buzzed around you enough to keep you awake. I know they are not conscious and not actively try to be nuisance, but still, I do not pretend to be fully rational either…

            Considering this, I do appreciate the effort but do not agree with the author classification of animal value. Contrary to relative intelligence were I think some ranking is theoretically possible, the animal value (in term of empathy worthiness) will be highly subjective and dependent on personal experience/activity. I doubt an old-style farmer and a city boy would agree about how pigs are worthy of empathy (especially if the city boy have a pet pig).

        • Le Maistre Chat says:

          Not sure for the original poster, but for me they are very different:
          – I was a children, I never was a cow.

          So if I want people to empathize with animals, get them to believe in reincarnation?

      • Nicholas says:

        Judging by our federal debt, “why would we, not being children, care what children want?” seems to be a shockingly widespread attitude.

      • Prismatic says:

        When I ask myself the question, “Why should I, not being Unirt, care what Unirt wants?” it’s pretty easy to answer. Unirt is a human (I’m assuming you aren’t an unprecedently advanced chatbot) just like me, and I want other humans to treat me with respect and charity. I’ve found the best way to do this is to be respectful and charitable of others unless I have a compelling reason not to. Our mutual and often tacit agreement to go, “let’s not be dicks to each other” is a backbone of human society which has (with some steps between) presented me with many nice things like computers, vaccines and professional wrestling.

        Likewise, I care about what people in Eritrea want to some extent despite not being Eritrean, never visiting Eritrea, and never having met or spoken to an Eritrean. This is because, like me, Eritreans are human and the best possible world would be a one where no humans suffer. Human suffering tends to lead to more human suffering for many reasons, such as the creation of a cycle of revenge that can also adversely affect third parties.

        Now, when I try to apply this same logic to a cow, I run into difficulties. Cows are not human. They have some resemblance to humans in their behavior, but I see no world where human treatment of cows has any broader impact on either as a whole. If all cows in all farms were emancipated, given land with which to roam freely and enjoy their lives, there would be no difference to the cows. They would not be grateful or angry because they cannot achieve mass action. The individual cow’s experience on this planet would be more pleasant, certainly, but the individual human’s would be worse thanks to the loss of animal products from cows and the economic damage done to the many people whose livelihoods revolve around cows. Now, you can argue that the improvement to Generic Cow’s life is way bigger than the downgrade to Generic Human’s life- and I would agree- but then I have to ask myself, exactly what is the ratio of cow suffering to human suffering?

        The article above tries to answer that question numerically, but I don’t agree with their numbers for the simple reason that I think animal suffering is profoundly less important to humans than human suffering is. A chicken’s suffering is not 10% as important as a human’s, at least it isn’t to me. Neither is a cow’s. Only animal who I believe can provide humans with something of significant value besides the fruits of their body- primates, cetaceans, elephants, the more popular pet species- have more valuable suffering for the (admittedly tautological) reason that we’ve agreed their suffering is more meaningful.

        tl;dr as someone who values the experience of humans significantly above the experiences of non-humans, I find meat-eating to be morally justifiable. I also think the above post, while interesting and well-researched, misses the simple notion that someone can support eating meat but not support factory farming.

        • Unirt says:

          A chicken’s suffering is not 10% as important as a human’s, at least it isn’t to me.

          And this is why the authors included their adjustable model so that you can adjust these assumptions for your own decision making.

          Reading this I notice that, in justifying why you care about people, you mention the possible gains to your wellbeing for being good to other people, both me and Eritreans. This differs a lot from my experience: I try to be good to Prismatic, as well as Eritreans, because I want you all to be happy. 🙂 The evo-psych reason for me feeling this way is certainly selfish, but I feel like this nevertheless. Whenever I think about Prismatic, or an Eritrean, or a cow, I automatically empathize with them, I feel like I was them for a moment and therefore I want them to feel good. I suggest you maybe feel something similar to your kids or other close relatives? There are probably inborn differences in who is included in one’s Circle of Empathy and this can include anything from “just me” to “all sentient beings”. So the cow question makes sense to anyone whose empathy doesn’t stretch to include cows.

          And I’m not trying to say that I’m a better person. I do, in fact, eat animals, even though I feel their pain; shouldn’t this make me a worse person compared to those who just can’t feel it?

        • Dacyn says:

          It sounds like you think about caring for other people entirely in transactional terms, or else in terms of secondary consequences. That is… disturbing.

          • Gurkenglas says:

            Depending on his decision theory, it is merely a more conscious variant of evolutions-based empathy.

          • Dacyn says:

            Yeah, but… part of the point of humanity is that evolution’s values are not our values. I know our values came from evolution, but I don’t want them to explicitly serve evolutionary goals.

          • Ghillie Dhu says:

            If one subpopulation’s values align with evolutionary goals and another’s does not, the former will supplant the latter; advocating against alignment is advocating for extinction.

          • dionisos says:

            If one subpopulation’s values align with evolutionary goals and another’s does not, the former will supplant the latter;

            This is false, it only mean somewhat less fitness, not at all extinction.
            Particularly when you take into account that :
            – When you know fitness is important to meet your end-goals you can take that into account. You can maximize it almost as much as those who have it as a end-goal itself. (so other considerations will matter much more)
            – Our rational decision making is only a very small part of our decision making. (akrasia)

            But much more importantly, it would be completely irrational to change our end-goals just because it’s harder to satisfy them.
            Otherwise I would just choose my goal to be maximization of entropy, or other instant win considerations.

            advocating against alignment is advocating for extinction.

            This is strongly false.
            Do you think extinction of the kind of people who follow fitness unaligned goals will favor those goals ?
            If no, then fitness unaligned goals aren’t advocating for extinction at all. (except if the explicitly do, and even if they explicitly do they would advocate for spread before being able to do it)

      • AlexanderTheGrand says:

        Or more generally, “why would we, not being anyone but ourselves, care what anybody but ourselves wants?”

    • Alex M says:

      I agree, it seems like the authors both share a moral foundation that I disagree with. This statement in particular makes the difference clear:

      Of particular relevance is the conscious experience of pain and suffering, which we regard as morally undesirable when it occurs in ourselves or others.

      Actually, I don’t view pain and suffering as universally undesirable, only in myself and those I consider my friends or allies. Pain and suffering are desirable in those I consider my enemies, because negative stimuli delivered in response to bad behavior encourage the enemies in question to change their behavior. This is evo-psych 101; I’m shocked that both of the authors could miss something so basic. Human beings literally EVOLVED to enjoy inflicting suffering on our enemies: this is why lynch mobs were so popular in medieval times, recent history, and now on Twitter.

      I suppose it’s ridiculous to categorize animals into “enemies” or “allies” but we can roughly extrapolate how we should feel about them by asking the question “If our positions were reversed, how would they treat us?” I’m pretty sure that most animals would feel no qualms about eating us for nutrition (though non-felines would probably be opposed to unnecessary suffering) so I think we should treat them the same way. Just my own two cents.

      TL;DR: Do unto others as they would do unto you. In game-theory terms, this optimizes your evolutionary fitness by allowing you to find effective synergies with other cooperators while pointedly discouraging assholes from betraying you.

      • Dacyn says:

        Most people are neutral, not enemies or allies. How do you feel about their pain and suffering?

        I don’t really see what it would mean for animals’ and humans’ position to be reversed. They don’t have the mental architecture to be able to run a factory farming system.

        • Aapje says:

          There are animals with a omnivore diet. When they hunt and kill animals, the suffering typically seems fairly extensive. Regularly, the animals ‘play’ with prey, torturing them. These animals don’t seem to stop hunting when sufficient vegetarian/vegan sources are available.

          • Dacyn says:

            Well, I’m not the game theory person but it seems like the scenario you describe has little game-theoretic relevance, since they aren’t hurting us (and they can’t, really).

            From my own perspective, it is an interesting fact but probably just indicates that those animals don’t have the capacity for the kind of empathy that would lead to taking action to avoid hurting others. I would be more worried if they had such a capacity, but chose not to use it. But on a species-wide level, I don’t think it’s even possible for such a dichotomy to hold.

        • Alex M says:

          Most people are neutral, not enemies or allies. How do you feel about their pain and suffering?

          In term of my ethical calculus, those people have a zero value: neither positive or negative. If they’re not with me but they’re also not opposed to me, I don’t care about their well-being one way or the other. I admit to having a mild (irrational) bias towards helping people, so in practice I suppose I would calculate their value at 0.00000001 utilons (where my average friends/allies lives are valued at 1 utilon, and average enemy lives are valued at -1 utilon) instead of a flat out zero value. In other words, I would generally try to be nice and helpful to neutral people, but if their well-being interfered in the slightest amount with my well-being or the well-being of my allies, or if injuring neutral people was necessary to successfully strike at my enemies, then I’d be OK with them becoming collateral damage.

          I don’t really see what it would mean for animals’ and humans’ position to be reversed. They don’t have the mental architecture to be able to run a factory farming system.

          Ask yourself the following question: if cows were smart enough to run a factory farming system, and if they were capable of deriving nourishment from humans, would they do so? Evolutionary psychology suggests that the answer is yes, so I believe that we have every moral justification to do the same to them.

          Another good follow up question is: would the cows farm us in a way that caused suffering? I believe that the answer to that question is no, since it is unusual to see cows display gratuitous cruelty. Therefore under my ethical system “Do unto others as they would do unto you” it’s OK to eat cattle, but not to farm them in a way that causes unnecessary pain or suffering. Their intelligence or alleged “consciousness” has no bearing on the ethics of the situation, as far as I am concerned. An AI algorithm may not technically be “conscious” or even alive according to the silly definitions of qualia, but if it is friendly to me then obviously I value its existence more than the existence of some Twitter rando.

          Hope that clarifies things! 🙂

          • Dacyn says:

            Regarding neutrals, I think Aapje said it well in the open thread:

            The common standard for ‘indifferent’ human behavior is actually a level of concern for others, which is below the level that we normally have for friends, colleagues and family, but above the level of a sociopath.

            I am having a hard time believing that you are a sociopath, though. So, as odd as it may seem, from my perspective the most logical conclusion is that you are mistaken about your own values.

            Ask yourself the following question: if cows were smart enough to run a factory farming system, and if they were capable of deriving nourishment from humans, would they do so?

            I think this is a fairly weird question, but anyway the only examples we have of animals that are intelligent enough to run a factory farming system are humans. So if they were that smart, I would expect them to behave roughly like humans. But that means if we decided not to factory farm, they would also decide not to do so, which means (according to your logic) that we would be justified in it. Anyway I don’t like this reasoning and think it is circular, but I think that that is just a sign that the original question is not a helpful one.

          • Alex M says:

            I am having a hard time believing that you are a sociopath, though. So, as odd as it may seem, from my perspective the most logical conclusion is that you are mistaken about your own values.

            Dacyn, “sociopath” is an insulting word, often used by inferior people who are unable to control their emotions with rational thought. Because such people tend to be narcissistic and ego-driven, they can’t accept that their inability to exercise control over their own emotions makes them inferior to others: instead, they rebrand their own emotion-control problems as a virtue (rather than the failing that it is) and label anyone who has more emotional self-control than them a “sociopath.” I don’t think that we’re going to have a very productive conversation if you insist on insulting me by labeling me that way. I assure you that I have emotions that I feel very strongly: I simply do not allow them to influence my value system.

            To explain my ethical system more clearly, the goal of a rationalist utilitarian is to maximize positive value for everybody in the universe. What maximizes positive value? From a game theory perspective, life is a Prisoners Dilemma. Positive value (peak utilitarianism) is maximized when every single conscious being in the universe is a cooperator rather than a defector. Therefore the goal of a truly effective altruist (who is logically consistent) is to convert defectors to cooperators.

            This can be done most effectively by the following strategy:

            1) Being a cooperator yourself (Self-evident)
            2) Rewarding those who cooperate with you (because that improves their reproductive success, allowing “good” cooperative behavior to culturally and genetically replicate)
            3) Punishing those who fail to cooperate with you (because that diminishes their reproductive success, making it more likely that their “evil” behavior will die out).

            In summary, good people have positive value, and we need more of them. Evil people have negative value, and we need fewer of them. In other words, if you embody the values of cooperation (as I try to do) then anybody who works with you cooperatively is good and anybody who works against you is evil. In rationalist terminology, by having a logically consistent value system and striving to embody the values that I seek to spread, my utilons and hedons are effectively the same thing: what’s good for me is also good for the universe (so anybody who works with me is good), and what’s bad for me is also bad for the universe (so anybody who opposes me is evil). I’m not mistaken about my values, it’s simply that I’m rational enough to have a value system that is logically consistent. Most utilitarians are logically inconsistent because they fail to recognize that not all lives are of equal value, even though that ought to be the most self-evident thing in the world. Does Mother Theresa’s life have the same moral value as the life of a child abuser? Obviously not. The world is a better place when Mother Theresa is alive and the child abuser is dead. Just because you may be too gutless to personally kill the child abuser doesn’t change that fact.

            I suspect that this logical inconsistency is a product of cultural brainwashing (the whole “turn the other cheek” mentality that in part stems from the Catholic church). From a game-theory perspective, “turning the other cheek” is an absolutely TERRIBLE principle to encourage better behavior from defectors – in fact, it makes society shittier by rewarding negative behavior. I don’t buy into this cultural brainwashing (because I want to build a positive uplifting society rather than a shitty one), and if you genuinely want to be a good person, you should strive to overcome this cultural brainwashing as well. Also, I recommend not calling people sociopaths just because you disagree with them. 😉

          • Max Chaplin says:

            > I recommend not calling people sociopaths just because you disagree with them.

            You gave most people in the world a moral weight lower than that of some particularly round pebbles. I’m not qualified to diagnose people with sociopathy so I’m not going to call you one, but this is exactly the sort of thing the popular caricature of a sociopath would say.

            Back to meat – if you’re comfortable basing your moral view regarding meat eating on a counterfactual prisoner’s dilemma, why not treat the efforts of ending factory farming as an attempt at making a values handshake with hypothetical cattle/chicken civilizations in return for them ending factory farming of humans?

          • Dacyn says:

            @Alex M:
            First of all, I did not call you a sociopath. I said some things which you combined with your own beliefs and inferred that I would call you a sociopath if I had more information. It’s possible the distinction doesn’t matter that much in this context but I want to make it clear that my intention was not to insult you.

            By contrast, your first paragraph gives the strong impression you are trying to label me as “inferior”, which is a word certainly used as an insult (in my opinion even more so than “sociopath”). Ordinarily I might excuse this by saying that since you perceived me as insulting you, you felt angry and therefore insulted me. However, you have just made a point of associating such “emotion-control problems” with inferiority, so you may not wish them associated with you. So perhaps by your own logic I should take your insult to heart.

            In point of fact I do not yet feel particularly insulted, but the fact that this is the third paragraph of meta that I felt I needed to say in order to properly respond to you, is not a good sign. So I would not be particularly surprised if either one of us ends up tapping out.

            Regarding the content of your comment: It seems like you are confusing the concepts of “ally” and “cooperator”. Most people are neutral to you in the sense of common language, but they still appear to be cooperators in the game theory sense that you describe.

            But more importantly: the thing that defines someone is the why of their ethics. And I gather yours is described by:

            the goal of a rationalist utilitarian is to maximize positive value for everybody in the universe

            This is in stark contrast to your earlier comments in this thread, where you said

            Actually, I don’t view pain and suffering as universally undesirable, only in myself and those I consider my friends or allies.

            What happened to “everybody in the universe”? Or is it that causing pain and suffering to your enemies actually leads to more positive things overall, enough to outweigh the pain and suffering?

            I don’t see how causing cows to suffer helps anyone else, except in the obvious sense that we get meat. And the whole point of this ACC was to try to weigh that consideration versus the suffering of the cows (and other costs/benefits).

          • Andaro says:

            Alex M, you should drop the utilitarianism label. Several of your points are inconsistent. You can make them more consistent by getting rid of the utilitarianism. I personally am a self-based reciprocator; I agree that people who benefit me have positive value, while people who harm me have negative value. I also don’t care either way about neutral people (people who neither harm nor benefit me). But this is inconsistent with utilitarianism, and I don’t use the label. It was a big flaw in this ACC that utilitarianism was taken for granted when instead this was the most questionable part. (Also you might be mistaken about the goodness of Mother Theresa’s choices.)

            Max Chaplin, why should we make value-handshakes with merely hypothetical civilizations? The whole reason we don’t have to care about cow or chicken welfare (outside of empathy management etc.) is that they don’t have intelligence and they’re not people; therefore they’re not part of the game theory of reciprocity.

      • Max Chaplin says:

        The enlightened view of punishment accepts that statement you’ve quoted as a baseline and adds exceptions with some reservations to wrongdoers with the stated justification of preventing even larger suffering. This is why the Geneva Conventions are a thing and prisoners have rights too – a punishment is an addition to the undesirability of suffering, not a replacement. If you don’t share these values then the article wouldn’t change your mind anyway and hence you’re not its target audience.

        Evo-psych is irrelevant here – it might tell you where your morals come from, but it’s not meant to instruct you which morals to adopt. Treating evo-psych as a moral guide is basically saying that the way things have worked until now is the way they must continue, but there is no justification for that.

        • Andaro says:

          In principle, we can agree that prisoners should have rights based on the possibility that we and all the people who’ve net-benefited us can become prisoners. We can have norms against torture, murder and similar human rights violations because we don’t want to be tortured, murdered etc., rather than deriving them from a universe-spanning utilitarianism that includes all kinds of non-reciprocators. It’s actually really hard to enforce such agreements, because punishing violations doesn’t necessarily pay; people can free-ride on the norm, as e.g. criminal torturers and murderers will do. It’s even more absurd to extend a norm that’s already hard and expensive to enforce to billions of cows and chickens who can’t understand a word we’re discussing.

          In principle, we could hope that vegans and animal rights activists are such great, altruistic people that they at least net-benefit us through their general superior benevolence and altruism, even though they want to make meat more expensive or outright ban it. The problem is that this often doesn’t pan out in practice. I’ve had too many discussions with vegans who think that if we don’t extend human rights to chickens, we don’t deserve human rights at all. This is of course a glaring defection. So not only are many animal rights advocates not helpful, they’re themselves free-riders on the human rights norm.

          • dionisos says:

            In principle, we can agree that prisoners should have rights based on the possibility that we and all the people who’ve net-benefited us can become prisoners.

            We can potentially agree but we don’t.
            The reason I agree prisoners should have rights is because it reduce suffering.

          • Andaro says:

            If that’s the main point, why not execute prisoners more often? In ways that are less painful than the average death? A dead prisoner can’t suffer and can’t cause suffering.

            If the utilitarians now object that this causes an opportunity cost in pleasure, why not focus on hedonium as your cause X? Or if hedonium isn’t realistic, the next best thing that is realistic?

            Yeah, it’s unattractive, and the reason it’s unattractive is because utilitarian rationalizations don’t map to people’s true motivations.

          • dionisos says:

            If that’s the main point, why not execute prisoners more often

            Because doing it will create a lot of psychological suffering, for the prisoners and also for society at large.
            It will also probably create a lot of other problems. (like political conflicts about it)

            Globally I don’t believe there will be less suffering inside a society doing it.

            why not focus on hedonium as your cause X

            I would be very happy with hedonium, but I don’t believe we are near of being able to do it. (particularly because we don’t understand consciousness well).

            Yeah, it’s unattractive, and the reason it’s unattractive is because utilitarian rationalizations don’t map to people’s true motivations.

            You are assuming too much. It is mostly a problem of strategy, you can’t durably reduce suffering by simple solutions though in 5mins, no more than you can become rich with these kind of solutions. (except for some exceptions)
            But if there were a possibility to do something like a hedonistic utilitronium shockwave, it would be perfect in my point of view.

          • Andaro says:

            No one said anything about 5min solutions. The point is that you’re not interested in actually working on the optimal solution space. Why doesn’t EA have a long-term strategy to research, invent and then give existence donations to hedonium? Because the utilitarianism isn’t their true motivation.

            “Because doing it will create a lot of psychological suffering, for the prisoners and also for society at large.”

            Is there any evidence of that? It seems very plausible to me that this is more than outweighed by the fact that these brains don’t exist and therefore suffer anymore + whatever reduction in violent crime you get from their nonexistence/deterrence effects. You may be right that it’s not politically tractable, but that’s also not clear to me, plus if that’s the truth, then how are prisoner’s rights generally tractable/neglected? Seems just inconsistent.

            I’m not saying you should be a utilitarian, btw. I think the correct answer is to drop the utilitariansim (and stop assuming that meat eaters should be expected to change their diet for it).

          • Aapje says:

            @Andaro

            People don’t exist in isolation. Most prisoners have family, friends, etc who get quite upset if they die, so killing a prisoner causes potentially persistent suffering in them.

          • Andaro says:

            Emotional suffering is hard to quantify, easy to fake and almost always assessed in a biased way. What about the ease of mind that a prisoner’s enemy feels when that prisoner is no more? It may well be perpetual!

            The point being that utilitarian narratives tend to be arbitrary.

          • dionisos says:

            No one said anything about 5min solutions.

            Yes, except me.
            My point is that a lot of counter-arguments to utilitarianism of the kind you give works ok in thought experiments but not so much in reality.
            I don’t believe most are really great ways to reduce suffering at all when we take everything into account, and you would probably not end-up with them if you was really trying to find how to reduce suffering.

            A lot of EA aren’t complete utilitarians (of any kind), so yes it isn’t their only motivation, but it isn’t hided in anyway.
            They still end-up taking reducing suffering into account and factually help reduce suffering, so I am all in favor of helping them.
            And I would also take their other values into account anytime as a way to cooperate with them, as long as they take mine into account.

            There are people focusing of stuffs like hedonium, I would say the hedonist imperative is pretty near of people doing this.
            And I think it is great.

            Is there any evidence of that?

            I have no strong evidence of it, but I don’t need strong evidence that it is negative to not try to do something difficult, only reason to think it isn’t strongly positive.
            (And thinking it is probably negative but without knowing much about it, is really enough reasons for not advocating for it)

            At most I am biased about not wanting to do it because it feels bad. (I indeed don’t feel super happy about it)
            But then I am only biased by my feelings, in the same way I could be for whatever goals I could have, it doesn’t mean I am wrong about what my goals are or lying about them.

            I’m not saying you should be a utilitarian, btw. I think the correct answer is to drop the utilitariansim (and stop assuming that meat eaters should be expected to change their diet for it).

            Yes I understand it, but even the alternative reason you gave to not kill prisoners, don’t work at all in my point of view.

            we can agree that prisoners should have rights based on the possibility that we and all the people who’ve net-benefited us can become prisoners

            Why would I identify with the criminals ? It is much more probable I become a victim of a criminal than I become a criminal myself.
            If I was caring only about what would favor myself or people I can identify with, and killing criminals deter crimes, I should be in favor of killing criminals.

          • Andaro says:

            “Why would I identify with the criminals ? It is much more probable I become a victim of a criminal than I become a criminal myself.”

            I wrote prisoners rather than criminals here because it’s possible to become a prisoner without first committing a crime. But even for criminals, governments can abuse their power to define what a crime is.

            “If I was caring only about what would favor myself or people I can identify with, and killing criminals deter crimes, I should be in favor of killing criminals.”

            I’m okay with killing some criminals. I see no rational reason to keep people like Anders Breivik or Marc Dutroux around.

          • dionisos says:

            And this reason to not kill prisoners works also in a utilitarian framework.

            If I want to reduce suffering, I probably don’t want people being deterred for doing anything which go against the power of the government to do whatever it like, or to permanently fear to be falsely accused and killed.

            I mean, I would not fear for myself about this except If I am in some authoritarian/dictatorship system. And then I would also really care about avoiding it in a utilitarian position.

    • dionisos says:

      Is this really more central than the question “Why would we, not being cows, care what cows want?”?

      Yes, because there is no rational reason to care about anything, animals, other humans, or even ourselves.

      It all come down to what we intrinsically value, and we don’t intrinsically value the same things.
      So the authors ended up on some kind of utilitarian positions, they, and a lot of others, can agree with, because their goal wasn’t to consider every possible ethical systems.

  8. eigenmoon says:

    Disclosure: will cook chicken right after pressing “Post Comment”.

    1. Pigs’ intelligence might be overrated. ESR writes: “a friend who grew up on a hog farm assures me that pigs bred for meat are stone-stupid; according to her, it’s only wild pigs I should be even marginally concerned about”.

    2. Hunting has completely different ethics: somewhat counter-intuitively, shooting an animal decreases overall animal suffering. The ethics of eating hunted meat is thus determined by whether you force somebody else to switch to farmed meat or not.

    Here’s why shooting an animal decreases suffering. Let’s say we have a stable rabbit population in equilibrium. If there are N rabbits, in a year about 10N new rabbits will be born and 0.1N rabbits will die peacefully from old age, and X rabbits will die horribly one way or another. Since the population is stable, we have N = N + 10N – 0.1N – X, so we get X = 9.9N.

    The important thing about this calculation is that the number of hunters is not in it. This means that hunting causes shifting the equilibrium towards reducing N while X/N stays constant, so counter to intuition hunting actually decreases the number of animals dying horribly.

    Maybe reducing the population is a bad thing. Fortunately, usually many species compete for the same resources, so hunting one species will increase the population of others. The competition dynamic is described by competitive Lotka-Volterra equations.My guess is that you just need to figure out the ratio of utility of different animals, and unless this ratio exactly matches the population ratio of the Lotka-Volterra equilibrium, you can increase the overall utility by hunting the animal that you disproportionately dislike.

    • HarmlessFrog says:

      Hunting also reduces the negative effect wild herbivores have on crops. Over here hunting lodges are responsible for wild animal damages to crops that happened on their turf, caused by animals they are allowed to hunt. Animals with hunting immunity get their damages compensated by the national treasury.

    • siwhyatt says:

      I think the fundamental flaw in this discussion is that it posits a false dichotomy: factory farming vs vegetarianism.

      I used to be a vegan, but after a lot of research and deliberation came to the conclusion that:

      1 The lives of free range animals are definitely worth living, probably “better” than most wild animals, and even many humans

      2 Hunting wild animals reduces animal suffering as noted above

      3 Well managed mixed agriculture makes better use of land and resources than 100% plant based agriculture.

      4 There is no good evidence that there are negative health effects from eating meat. Even with the Adventists, it’s not so simple to say the only difference was animal products. Any arbitrary dietary restriction results in improved health. Whether you cut out meat, gluten, carbs or ingredient X you end up cutting out pepperoni pizzas, burgers, etc.

      Yes, factory farming is horrific.

      Yes, we should probably greatly reduce our meat consumption and vary the species we eat from just 4.

      But framing the choice as factory meat or no meat I don’t think is useful.

      • ACC2019 Omnivore says:

        I take your point. There were multiple questions we could have considered investigating. Whichever one you were most interested in, I hoped the ground we covered touched on it. “Is the status quo preferable to one where a typical omnivore in the west switches to vegetarian?” is the precise framework we agreed on.

        Re. your particular points (speaking for myself)
        1. Agree strongly. We concluded that factory farmed cows definitely had positive lives (from their perspective; ignoring environment, health impact on humans, finance, etc). If what you care about is only animal lives, feel free to eat anything with living conditions at least as good as factory cows.
        2. Haven’t done any research, won’t comment because I don’t know how hunting affects long-term animal population. In farming the alternative is non-existence.
        3. This is a very complicated claim to evaluate, one on which we may not disagree, and hinges on definitions of ‘well-managed mixed agriculture’ and ‘better use of land and resources’ (do you care about soil integrity, greenhouse gas emissions, biodiversity, etc). Our conclusion is that given current magnitudes and existing implementation of animal farming no animal farming would be significantly better for the environment.
        4. There is a bunch of weak evidence that we found fairly compelling in aggregate, for the claim that a typical western omnivore diet is less healthy than a vegetarian diet. We are not claiming that the vegetarian diet is optimal, or even that we know what is.

        • thevoiceofthevoid says:

          Re: 4: Was there ever any proposed mechanism by which eliminating meat produced better health effects? This seems especially important in light of the list of nutrients for which the primary source is meat consumption. I would be more receptive to the observational studies if there were at least some causal explanation to back them up.

          • Randy M says:

            Yeah, that stood out to me. “Meat is more nutritious, but will kill you.” An attempt to resolve that might suggest a more optimal diet.

          • ACC2019 Omnivore says:

            No. Here we frankly ran into expertise constraints. It’s hard enough to figure out what’s healthy for you without any randomized controlled studies; evaluating the plausibility of various proposed explanations was beyond us. Going in I had a strong prior that meat was healthier because more nutritionally complete, and necessary for evolution of larger brain (turned out to be for caloric density and time opportunity cost of chewing raw food reasons, not nutrition, so we cut this section), but couldn’t substantiate. There was enough circumstantial evidence that typical vegetarian diets (in non-infants, in developed countries) are healthier that I was persuaded to change my mind.

          • thevoiceofthevoid says:

            @ACC2019 Omnivore

            My theory is that, in addition to the “anti-inductive” effect you mentioned: People in developed countries generally eat more calories than is optimally healthy, and going vegetarian generally reduces the number of calories you eat (since e.g. hamburgers and pepperoni pizza are now off-limits). I suspect that vegetarianism has similar effects to cutting out any other perceived source of “badness” from your diet (e.g. carb-free, sugar-free, or fat-free diets), or just counting and reducing calories in general. I don’t feel like trudging through the mire of nutritional research to prove or disprove this, but I’d be interested in the analysis if anyone else is.

            If true, this would explain the benefits of vegetarianism all the studies find and imply that going veg*n probably is better for you than just eating all the crap you usually eat, since reducing “junk food” intake and/or cutting calories is pretty uncontroversially good for your health. However, it would also imply that if your main concern is your personal health, cutting out meat probably isn’t the best way to go about it compared to other diets, given the various nutrients best supplied by meat.

            So I think your conclusion may not be wrong, per se, but is somewhat misleading in that it implies that meat causes health issues directly in a way I don’t think is sufficiently supported by the data.

          • DeWitt says:

            Why would you just assume becoming vegetarian also implies cutting out junk food? Really, why? I went vegetarian for a few years in my teens, and while that meant no burgers or kebab I could have whatever soda and french fries I wanted.

          • thevoiceofthevoid says:

            @DeWitt

            Obviously vegetarianism doesn’t necessarily entail cutting out all junk food, but there are certain common categories it necessarily eliminates (like burgers and fried chicken). Also, while you can still have soda and french fries, you’re a lot less likely to want to go to McDonalds when those are the only things you can eat there.

      • RalMirrorAd says:

        Could you substantiate #3? what research lead you to this conclusion?

        • siwhyatt says:

          Ruminants can be raised on pasture land which is not suitable for growing crops that can be eaten directly by humans and pigs and deer etc can be raised in forests, thus significantly increasing the land available for farming over a vegan diet. Both scenarios help with carbon sequestration.

          Animals can also be fed a lot of “waste” products from plant based agriculture – the inedible stalks of corn and soy, residues left of cereals after processing for beer and biofuels, vegetable peelings, etc.

          The bones, blood and manure of animals can be used as fertilizer, and leather, gelatin and a host of other products come from livestock also.

          Stop farming animals and we lose a great deal of land for food production, lose a very efficient way to “recycle” the byproducts of plant based agriculture, and lose a way to make pasture land and forest profitable.

          That said, of course, if we were to raise livestock on 100% natural forage plus waste vegetable matter in ethical conditions it would require a massive reduction in numbers.

          Apologies, I’m not great at saving references, but here are a couple of studies related:

          http://www.jswconline.org/content/71/2/156.full.pdf
          (Role of grass-fed ruminants in sequestering carbon)
          https://www.elementascience.org/articles/10.12952/journal.elementa.000116/
          (This study finds lacto-ovo vegetarianism to make the most effective use of land but doesn’t take into account the possibility of 100% pasture raised ruminants, nor forest raised animals or wild game)

      • Froolow says:

        (I’m the vegetarian collaborator)

        Just to add to my partner’s comments – I think framing the choice as factory meat or no meat is probably fairest from the perspective of a typical SSC reader (which was the original framing of the collaboration). Our research suggested unless you can personally trace the source of the meat – for example because you hunted it yourself – you should assume you are eating factory farmed meat, especially if you are eating poultry. The typical SSC reader will eat a lot of poultry and will not hunt, so the characterisation seems fair to me.

        I take your point that there might be hypothetical industrial meat production systems that produce net positive results, but trying to answer what meat production ‘should’ look like was a much bigger task than we felt equipped to handle

        • Blueberry pie says:

          A very minor thing – in EU it is usually reasonably easy to trace most of meat you buy directly, due to mandatory labelling (but restaurants usually don’t provide that information)

    • melolontha says:

      1. Pigs’ intelligence might be overrated. ESR writes: “a friend who grew up on a hog farm assures me that pigs bred for meat are stone-stupid; according to her, it’s only wild pigs I should be even marginally concerned about”.

      It might be overrated because… a famous computer guy has a friend who thinks so? I guess that’s not literally zero evidence, but it’s got to be pretty close.

      • Taleuntum says:

        I agree, it is a very strange piece of evidence to bring up. While reading I expected the quotation to end with “She almost always beats them in chess.” or some other joke.

      • eigenmoon says:

        You’re completely correct.

        I just find it somewhat frustrating that people talk about intelligence of [species] without specifying where did they get the specimens. For example, house cats are less intelligent than feral ones (citation missing).

        • Nornagest says:

          I think I’d expect house cats to be more intelligent than feral ones — they’re the same species, and house cats get considerably better nutrition. Ferals might be selected for intelligence in ways that domestic cats aren’t, but spaying/neutering has been common for less than a hundred years, most domestic cats aren’t pedigreed, and a good chunk of them come from feral stock captured for use as pets, so there’s lots of genetic interchange between the populations.

          On the other hand, housecats might be dumber than their African wildcat ancestors, and dogs might be dumber than wolves.

          • MNH says:

            Pretty curious about dogs vs wolves. On one hand, there’s some evidence we did something kind of William’s-syndrome-y to them, but on the other hand it seems like some breeds got some serious positive selection on intelligence (e.g. Border Collies)

      • pepys says:

        That kind of evidence seems to be much of this comment section. Some guy who eats no plants citing a Youtube video to make canola oil seem bad (the video says canola oil is healthy), a friend of whoever ESR claiming that domestic pigs are dumb, someone who thinks we should purposely cause animals pain because they would eat us if the positions were reversed (this is apparently evo-psych 101). It just goes on and on.

        This is the worst SSC comment section I’ve ever seen, and I’m not even a big fan of the article.

        • dionisos says:

          Yes :-/

          someone who thinks we should purposely cause animals pain

          Not the first time I see it in this blog.
          I just hope both comments was from the same person.

          • markus says:

            +1

            But on the other hand. The extra mental power to log in and write this came from somewhere. A inborn willingness to punish those harboring such an evil, ill-argued and laughable view?

    • gkai says:

      pigs bred for meat are stone-stupid

      It’s not only pigs, domesticated animals tends to be more stupid than their wild counterpart….

      They are selected for this…well, not selected for stupidity but for manageability (and also lack the constraints on the wild on being intelligent enough to survive).
      So domestic animals tends to be much more placid, which often means also stupid.

      I’d argue that this is an important factor when considering animal happiness: a wolf would not be happy sharing a house with a human family (and he would makes you aware of it’s unhappiness – dangerously so), but most species of dogs are, while they would be unhappy (and soon dead) in the wild. Same for cow vs auroch, pigs vs boar, probaly chicken v.s. red junglefowl, depending on how much selection has changed wild form…..
      My rule of thumb about domesticated animals happiness is that if they do not actively try to escape, they are happy (or so dumb i don’t care).
      Industrial farming a little bit new, but, I think that if one cared enough, we could select animals happy to live in those condition (happy = not trying to escape, so selection may be to makes them extremely stupid, even more so than they are)….
      But one does not care enough, for ultra-placid chickens to be selected they would to reduce the cost of the infrastructure due to their happiness, so it remains theoretical…

      Still I’d like to know how people choosing vegetarianism due to concern about animal well being would react to a solution involving selecting chicken “happy” to live in industrial farming condition….

      • Unirt says:

        Not trying to escape is frequently used as a measure of depression in rodents in behavioural studies (normal mice run around looking for an exit, depressed ones cower in the corner waiting for death or something). I don’t therefore think it would make a good indicator of happiness. Also, it should be obvious, if you can see no way out and have never seen a better place, it may not occur to you to try to escape.

        • gkai says:

          Which illustrate how slippy the happiness concept is, even for humans. A lot of study shows that people report happiness level only weakly related to observable factors, instead it seems base happiness is largely innate. Change (especially transient changes) in happiness level are strongly affected by externals, but base level not so much…except maybe extreme externals????

          Philosophical question: living the same exact life, would you be more happy knowing there is a happier alternative but you are somehow unable to get it, or ignoring this happier life was possible?. And what about if the difference is in the way you are prevented reaching greener pasture: passive agent or decision by an active agent?

    • onyomi says:

      pigs bred for meat are stone-stupid

      On the other hand I’ve seen the brain of a pig I assume was bred for meat in a Chinese kitchen before and… it’s really disturbingly large. If you had told me in a different context it was the brain of a 7-year old human I’d have had no cause to doubt you.

    • mwengler says:

      “a friend who grew up on a hog farm assures me that pigs bred for meat are stone-stupid”

      I’ve noticed that people who are against immigrants think immigrants are stupid or otherwise less valuable. People who are against racist or gender equality think the races or genders they are against are stupid or otherwise less valuable. How much should we weight the opinion on hog stupidity of someone who keeps hogs for slaughter?

      • eigenmoon says:

        Unfortunately you’ve given it a very CW-prone framing and I was very tempted to jump in explaining why I can’t take any progressive opinion seriously anymore.

        But your concern is quite valid. Scientists believe now that cows and sheep are much smarter than farmers thought (citation needed), so maybe farmers are wrong about pigs as well.

      • Thegnskald says:

        Causality may flow the other way.

        I grew up around agriculture. Cows are placid, farm chickens are both stupid and evil. Didn’t interact with pigs at all except seeing them at livestock shows, which really doesn’t count.

        I don’t care about the suffering of farm chickens; I liken it to cloning Hitler (with memories and continuity intact) over and over and torturing him, while providing a benefit to the rest of society. Show chickens are somewhat smarter and more pleasant, albeit not a lot; maybe I’d care if they were being factory farmed?

        Cows? Eh. I guess I object to particularly inhumane treatment, but it is hard to clear the bar of how they would die in nature. (Hell, how a calf on a neighboring farm died – by natural causes, and I wi say no more – is nightmare fuel.)

        I don’t seek out pork, but if I did, I guess I’d update in favor of not eating it. Since I don’t like the flavor anyways, just doesn’t feel pertinent.

        • gkai says:

          Care to elaborate why chicken are evil? Apart from being distant relatives of T-Rex?

          I have interacted with chickens in my young age (from time to time, had friends who kept them for eggs and meat), and certainly agree they are stupid.
          Not judging them compared to mammals (being one, I may have favorable bias), but from other birds. Crow and family are a pleasure to observe. Chicken are not, the only fun I had was seeing how their head is stabilized, a biological steadycam 😉

          But they didn’t strike me as particularly evil….

          • Thegnskald says:

            Not chickens, the specific little bastards they factory farm. What I think of as “farm chickens”. They’re vicious, petty, cruel, and way, way more stupid than normal chickens.

            Like, if you let more than one roam around, odds are that sooner or later the smallest is going to be plucked naked, and have constant injuries from pecking. If it doesn’t starve or dehydrate because the other chickens won’t let it near the food or water.

            Google “chicken bullying”. That’s normal chickens. Double the aggressiveness and the stupidity for the chickens bred for factory farming.

            (Also, cannibalism isn’t limited to factory farming, it just gets intensely bad there because it is a learned behavior, so spreads like an infection.)

          • onyomi says:

            It’s interesting the chickens they bred to be factory farmed became not just stupid but aggressive and evil. That’s not what I would have expected. Superficially I would have expected factory farmed animals to be bred for docility.

            But I guess the farmers are more concerned about the meat so maybe the factory farm chickens are bred to be… buff chickens?

            Maybe factory farm chickens are the chicken equivalent of this? But I’m not sure whether that should make me feel less bad about eating them and/or worse that humans brought such creatures into existence in the first place?

          • gkai says:

            I would have thought that too, as it is what usually happen: domestic variants far more docile than wild one…
            But thinking more about it, factory farming means no (or little) human handling of the chicken, so the only selection target was probably grow as fast as possible. Logical strategy is to keep the fastest growers for next generation, easy: reproduce the bigger ones at slaughter time.
            Problem is that a metabolism tuned for fast growth is not the only way to become big fast, eating more than your neighbors and/or hurting and/or killing them makes you the fastest grower too, especially in relative term…
            Seems logical and could be corrected (probably quite fast) by selection, removing the nastier one even if they are the fastest growers…
            Multi-criterium selection is harder though, you have to monitors your chicken much more

      • Nornagest says:

        Probably more than someone who’s never seen a hog that wasn’t either in a petting zoo, or sectioned into little pieces and wrapped in plastic.

      • JohnBuridan says:

        My wife’s family does some pig farming. She says they (the pigs not her family) are quite smart. They can solve some puzzles, and notice when other pigs are being taken away to be killed vs. groomed or shown at the fair.

        In my own interaction with a pig, he overpowered me when I opened the gate and escaped. I must admit, at that moment, I felt the eternal footman hold my coat and snicker, and in short, I was afraid. The pig knocked me unconscious, dragged me into the pig house, and put me through a little procedure; it didn’t last long. It just gave me henceforth a little tick where once in a while I exclaim, “Four legs good! Two legs baaaad!”

    • DragonMilk says:

      On 1, Pig intelligence is definitely not overrated. I base this on my father dissecting dozens of pig brains and his experience, as well as berating me for being at times no smarter than a pig!

    • Michael Watts says:

      Here’s why shooting an animal decreases suffering. Let’s say we have a stable rabbit population in equilibrium.

      The assumption that a wild population is in stable equilibrium is not justified.

      • eigenmoon says:

        Lotka-Volterra equations have cyclic solutions, so you could take an average over a cycle to get a simple calculation similar to the calculation at equilibrium. Also the hunters might decide to hunt unevenly during various phases of the cycle in order to bring the system closer to its equilibrium.

        • Michael Watts says:

          Also the hunters might decide to hunt unevenly during various phases of the cycle in order to bring the system closer to its equilibrium.

          Isn’t the hunter deciding to hunt unevenly during different phases of the cycle the main thing keeping the system from reaching equilibrium?

          • eigenmoon says:

            As I’ve said, Lotka-Volterra equations have cyclic solutions. This means the system isn’t moving towards equilibrium by itself.

            If the hunters work as a team, they can make the system spiral towards equilibrium, just as you can control the amplitude of a swing by giving it a gentle push at every swing.

            If the hunters don’t coordinate, then they’re pretty much part of the system. And the system isn’t moving towards equilibrium by itself.

  9. Surfraider31 says:

    The habituation section misses something important which counts considerably against eating meat. It’s true that people’s happiness and sensations habituate, but we generally don’t habituate to things causing immediate pain, and I’m pretty sure this is one of the big factors that make people much more likely to want euthanasia (which might have been a better literature to look at than the suicide one). Low level pain like back pain we don’t want to trade many QALYs for, but the exchange rate increases pretty quickly as the pain goes up. Another thing we tend not to habituate well too is boredom – there’s those psych studies showing people knowingly inflict pain on themselves to escape it. It seems that many factory farmed animals experience moderate to strong chronic pain, and those that don’t like the pigs in crates experience boredom, and we can’t assume they habituate to these things by looking at how humans habituate to paraplegia. They aren’t analogous. Doesn’t change the overall conclusion but this gives us reason to think the vegetarian’s starting estimate is more accurate.

  10. statsman says:

    I only read the section on health as this is the bit I was interested in.

    Disclosure: I have a lot of nutrition related health issues and have read multiple cubic meters of relevant academic literature. I also studied the relevant statistics intensively.

    My conclusion is that the nutrition section is superficial, worthless and so bad it is not even wrong. Weak epidemiological / observational evidence, studies by motivated researchers, studies that ignore important information, ignoring important studies etc etc.

    • HarmlessFrog says:

      My conclusion is that the nutrition section is superficial, worthless and so bad it is not even wrong. Weak epidemiological / observational evidence, studies by motivated researchers, studies that ignore important information, ignoring important studies etc etc.

      That’s the field of nutrition for you. This would be funny, if it weren’t so sad. It’s even sadder that the folks in charge of nutritional recommendations/guidelines apparently don’t have much more than this superficial understanding of what they’re doing.

  11. Bellum Gallicum says:

    Famers are overlooked as is human growth and performance, which is the purpose of food consumption.

    • Unirt says:

      Farmers would switch to growing more plants if everyone would suddenly turn vegan, and all would be happy? If there was less work for conventional farmers as a result, more farmers would go for luxurious niche markets like organic products, for people would have more money to spend. Am I missing something?

      Human performance is accounted for by the health discussion, even though the data is low quality, as the authors agree. What other kind of performance do you mean? Even bodybuilders can apparently subsist on vegan diets.

      • caryatis says:

        You’re missing the economic and personal cost of switching from one career to another. If society eliminated your industry, you could find another job eventually, but it would be highly costly. Same thing for farmers.

        • siwhyatt says:

          It also misses the fact that if you have been farming sheep, aka wooly cows, on the cold, wet, windy highland slopes somewhere in the northern latitudes, you can’t just switch to farming avocados ‘cos that’s what the kids want nowadays…

  12. silver_swift says:

    To give a counterpoint to the vibe of negativity in the rest of the comments: I very much liked this AC. I feel it is a little more one sided than it could be, but it makes a lot of interesting points and I definitely enjoyed reading it.

    Animals we intuit as conscious are less likely to exhibit ‘glitching’ behavior indicative of being a non-conscious rule-following automaton.

    It’s a good thing human brains don’t have any glitches then.

    • onyomi says:

      Yeah, I am not a vegetarian but I think they did a great job.

    • eccdogg says:

      I also thought this was well done. I am largely a vegetarian that occasionally eats seafood.

      This article probably moved towards adding some (non factory farmed) meat to my diet.

    • inhibition-stabilized says:

      Same here. There are obviously a lot of points up for debate and many other angles the question could be approached from, as the rest of the comments show, but overall I think they did a fairly reasonable job attacking a really complex question. I especially appreciated the inclusion of the spreadsheet used to calculate the results and the sensitivity analysis at the end.

    • Jefferson says:

      I also agree with this – it seems that there are definitely more things they could have covered, but they had limited time and I think what did get covered was done thoughtfully.

    • Thegnskald says:

      Omnivore.

      I’ll also say I approve. I’d say the impression of lack of balance isn’t accurate, but is understandable. Fundamentally, the eating meat side has “nothing to prove”, and is in a somewhat defensive posture here, fundamentally having to argue against multiple lines of criticism of the practice. The lines of attack that are available (the classism and first-world-blinders that vegetarian proponents tend towards) don’t seem relevant here, possibly because overall, I just don’t feel like poor people who eat meat are being denigrated.

    • salvorhardin says:

      A lot of the one-sidedness seems to me to be because the omnivore seems way way more rationalist/EA-ish in their approach to morality than the vast majority of omnivores probably are. So for instance they let pass as conservative a 10% suffering weight for chickens which seems obviously at least two orders of magnitude high to me, and in their variant assumptions section, rather than looking at a more realistically/strongly speciesist weighting they only look at the even more ridiculous “no speciesism” weighting. On the other side of the argument, the Repugnant Conclusion-ish “eating beef is morally fine as long as the cows’ lives are not so bad that they would seriously prefer nonexistence” is also going to be very counterintuitive reasoning to most people, including omnivores.

      But we are, tautologically, selecting AC participants from among the sort of people who read SSC and want to be AC participants, who are probably so much more rationalist/EA-influenced than the general population as to make this bias unavoidable.

      • sty_silver says:

        I read this post as assuming that their “bias” is obviously a bad thing. But these are mostly factual questions, and the “bias” is either correct or not (or partially correct, but you get my point). You may believe it’s incorrect and that’s fine, but I don’t think you acknowledged that you may be wrong. This reads as if you think the question of how much chickens suffer is a matter of opinion rather than an unknown.

        • Said Achmiz says:

          How much I care about chickens (regardless of whether or not they “suffer”, regardless of whether or not it even makes any sense to speak of such animals “suffering”, etc.), is certainly a matter of my moral views, and certainly not a matter of any external “facts”.

          • sty_silver says:

            Only if you assume that moral realism is false, which you could also be wrong about. But the post I responded to didn’t make it sound like moral views were the crux.

          • Said Achmiz says:

            No, that assumption is not required.

      • Said Achmiz says:

        rather than looking at a more realistically/strongly speciesist weighting they only look at the even more ridiculous “no speciesism” weighting

        Yes, this jumped out at me as well. The one-sidedness of this AC was never quite so blatant as in that bit.

    • craftman says:

      I also enjoyed it if not just to give me a better framework within which to justify my meat-eating.

      Regarding the section on costs…my own personal calculus heavily downplays the health risks. I eat moderate amounts of meat and mostly exercise a bunch to live life as an objectively healthy 34 year old. I also put much less probability on chickens having a meaningful consciousness, which takes out another large chunk of the harm reported here.

    • Just want to pile onto this, as well. I eat meat*, but I really appreciated the wealth of data in the collaboration. While I was reading, I was gearing up to saying “I’d like to see your math, so I can see what happens if I plug in different values”, and then near the end of the collab that was exactly what was supplied, so I was pretty happy with it, overall. After all, there was a lot of eyeballing of potential values, but I felt it was fair to do so, especially giving people the option to eyeballing it differently to very tangible effects!

      I could probably say more about this, but I’ll leave it at a simple note of appreciation for now. 🙂 Thanks for the collab, Froolow and David!

      (* if anyone’s curious, I won’t be stopping in the near future. This isn’t because I’m not considerably swayed by this collab, which has significantly shifted some of my priors (chicken suffering being quite that bad, whereas cows have it comparatively good), but because the reason I’m eating it is (mild) nutritional phobia, initially set on by an actual deficiency. I do expect I will get better at this (I am basing this opinion on a different health-related fear that has declined over the years), but at the moment the idea of not eating meat is actually viscerally terrifying (in that I start to panic if I’ve not eaten any in a few days – being reasonably sure that I’m overreacting doesn’t dampen this effect, unfortunately). However, I am reasonably sure that the more the memory of my suffering fades, the less terrifying it will be, and eventually, I will venture back toward trying vegetarian diets. I certainly would like to.)

  13. dirtyid says:

    I think these net harm discussions are incomplete without also measuring and comparing the net harm of meat or vegetables production to both animals AND HUMANS. Compile a table equalizing sum labour hours required to produce a caloric unit of meat versus greens, factoring in stages of labour for animal feed production, and normalize that for injury + fatality stats, and extrapolate for amount of workers in each field. I’ve crunched rough numbers using limited datasets before and the conclusion is: meat production harms much less humans for equivalent caloric production.

    The reason is pretty straight forward, animal feed and animal meat production is highly automated and mechanized, wheres a outside of staples – that is the basis of both vegetarian and animal feed – many fruits and vegetables are extremely labour intensive to produce, espeically if the produce is imported from countries with low agricultural mechanization, you’re looking at 100-200 more labour to generate a kg of some produce compared to meats that are not beef. With mechanization it’s around 5x more labour. Normalize kg for actual calories (meat is more calorically dense), and the numbers are even more skewed. I think one of the comparison was that 1kg of hand picked apples has the equivalent labour of 50kgs of mechanically produced meat. Also beef is irredeemable.

    Extrapolate that there are magnitudes more farmers than meat production workers around the work, particularly in subsistence societies – billions versus millions, that agriculture labour in general is comparable or more dangerous in terms of injury and fatalities, affects younger workers (literally children too young to work in meat plants), the fact that it happens outside in large fields that are difficult to regulate and the moral calculus is hundreds billions of dead animals versus non insignificant human misery. If we want to reduce human harm, we’re better off exporting factory farm produced meat fed with mechanically produced feed. Automation + mechanization i.e. robopickers is not an option in developing economies when it comes to agriculture processes for the foreseeble future. Subsistence farming is probably the 1st civilized profession, and it’s here to stay as long as people stay poor, or rather, until societies get meaningfully rich. Until then, the diet that harms the least humans is animal feed followed by a combination of animals + animal feed. Again, not cows. Anything out side of meagere vegetarian options, i.e. “luxury” vegetable variety, dramatically ramps up human harm.

    E: dug up old napkin math

    https://docs.google.com/spreadsheets/d/e/2PACX-1vTcZ09yvuwfa0qB3ePamvOCX0upgZUHBzJ3repz-kL5cDYEl-Ga5QqpnW-jimFRVIopbR8iomRHsaVv/pubhtml

    • This is a really interesting point (although I haven’t independently dug into this right now, but it’s food for thought that I hadn’t had before, and that alone is worth something to me right now). Thanks for adding it to the comments!

  14. When talking about the environmental impact of beef it’s important to distinguish land use that could be used for agriculture and land use which couldn’t. A lot of cow grazing happens on scrub land which doesn’t receive enough rain to be easily used for agriculture. This doesn’t make the feedlot stage of a cow’s life good, of course, but it seems like the USA could produce about half as much beef as it currently does without feedlots.

  15. Robert Jones says:

    This is much better than the last one. You have to respect a post where section 1.1 is “What is consciousness?”

    We’re not sure about chickens. We encourage you to read this overview of their behavior in full to convince yourself that their emotional and cognitive intelligence would group them with simple mammals if they had the same neural architecture.

    However, since in most parts of the human brain ‘intelligence’ does not correspond to ‘consciousness,’ and because chicken brains are a clump of neurons with a different evolutionary history and lacking the distinct layered and highly folded structure of the cerebrum, in the model we assume their likelihood of consciousness is 75%.

    The base case assigns a 75% chance that chickens are conscious, and this is a big assumption to which the model is highly sensitive.

    It would be good if the authors could give some indication of where the 75% figure has come from.

    I note that Marino’s affiliation is given as The Someone Project, The Kimmela Center for Animal Advocacy, and that the Someone Project aims “to increase awareness about the complex minds and lives of farmed animals and influence farm animal policy for the benefit of the animals themselves.” This suggests some risk that Marino may have predetermined her conclusion. How confident are the authors that her article reflects consensus?

    If you eat meat and cannot explicitly trace the source you are most likely eating factory farmed meat.

    What if I eat chicken which is labelled as free-range? Or organic? Are those chickens leading net-positive lives or not?

  16. NostalgiaForInfinity says:

    Woolly cows made me laugh.

    Generally I thought this was a pretty good read. It seems thorough given the varied literature to draw on, and the final presentation of results including sensitivity was good. Reasonable people can disagree on the conclusions / quality of the evidence presented (or omitted). But the article as a whole was good given the impossibility of providing an exhaustive literature review of about 6 fields.

  17. bwingrave says:

    What is “liter”? 2.2.1 includes the following sentence: “Slaughter chickens in the West are raised instead in a large ‘broiler’ shed covered with liter.”

  18. Nicholas says:

    And this is the absolute crux of this investigation; if animals suffer under current farming standards to the point of preferring non-existence then there is a moral burden on meat eaters to justify eating them.

    The proposition that suffering animals still derive some positive utility from their life, so better for them to exist in agony than not exist is presented as more or less a given, but it doesn’t seem straight forward to me. I haven’t read past the introductory paragraph yet, so maybe this is fleshed out further in, but it seems like the inflicting of any suffering should require a moral justification.

    Otherwise, I don’t see how you could avoid giving a pass to the whole slavery industry: so long as slaves didn’t commit suicide their lives must have been, on balance, worth living, thus requiring no moral justification for the suffering inflicted…? It seems like any chain of moral reasoning that can be applied to slavery and return a “morally okay” must be flawed.

    • J Mann says:

      The authors assume that the alternative to factory farming is that the factory farmed animals not exist. The alternative to slavery is presumably that the slaves are free. (Industries will still need workers, so there’s not much reason to believe that banning slavery reduces total population).

      if you had some extreme Omelas case – let’s suppose some situation where in the absence of slavery, human life on a given island would literally die out – you’d have a different ethical problem where a consequentialist might conclude slavery was justified, although many other ethical systems (including mine!) wouldn’t.

      • Nicholas says:

        Since slavery was practiced multi-generationally, the children born into slavery wouldn’t exist without slavery in exactly the same way chicks hatched into factory farms wouldn’t exist without factory farming.

        ….I hadn’t considered this at the population level, and may have to think about it further, but it doesn’t immediately change my moral intuition on the validity of the analogy.

        Edit: having reflected a bit more: both utility and suffering are experienced subjectivity by individuals, not by populations, and the calculus of life with suffering > nonexistence is derived subjectivity by an individual, not through a population. So total population numbers cannot, I think, abrogate the need to justify inflicting suffering under the framework proposed. Population may be used as a justification for inflicting suffering, as in your hypothetical, but not a reason to not need a justification for inflicting suffering.

        • J Mann says:

          My read of the author’s assessment is that they’re looking at total conscious life for the (with factory farming) and (without factory farming) comparison hypotheticals.

          Under their system, as I understand it, if a cow’s life is net positive, then more cows improve total welfare, even if it’s not as positive as it would be in a world where cows had more rights.

          In your hypothetical where the states (say) outlawed slavery in 1789 across the nation, you’re right that the slaves wouldn’t exist. Instead, presumably, there would be free people in Africa who had never been enslaved, and a higher population of free (or indentured) people living in America. Assuming those lives in the other timeline are more positive than the lives of the slaves, that’s a consequentialist improvement.

          Presumably, in the hypothetical where we don’t eat factory farm beef, there would be *some* additional offsetting conscious lives – since meat eating is relatively inefficient, switching away from it would leave some lands fallow where we now grow cattle feed. As a result, deer and small mammals (and their predators?) would be able to increase their populations, and it’s possible that the life of a wild deer, rabbit, or fox is more positive than a factory cow. I’m not sure how many of those lives there would be, or if the authors considered them.

          • Nicholas says:

            Yes, that is my interpretation of the author’s theory as well, yet I don’t find it convincing.

            I haven’t studies the subject, but my assumption is that the incentives and economics of slavery would indeed lead to dramatic increases in total population over unfettered tribal life. However, even if I’m wrong in that assumption it seems bizarre and perverse to me to argue that the greatest moral failing of slavery wasn’t the horrors inflicted on individuals, but rather that slave owners neglected to force their victims to have enough offspring with tiny positive utility to offset the huge negative utility slavery was inflicting on the population neutral set of slaves.

            Furthermore, I am not convinced that you can snapshot at a particular moment in a life, and if that being isn’t actively suicidal, conclude their existence must have positive utility. How much of the will to live amidst abject suffering is merely hope of one day escaping? Taking utility deficits year on year in the hopes of a utility payoff later?

            We need the opposite of the veil of ignorance: a veil of perfect information that can show an unborn entity the entirety of their existence, and only if they understand the totality of their suffering, with no happy ending, and *choose* to exist in the first place would I accept that as a positive utility life for a factory chicken (or slave).

            Even if a life is unambiguously positive utility, I’m not convinced that alone means no moral justification is necessary before inflicting suffering.

            There are so many levels that this logic is at least questionable that I don’t think hand waving it away in two sentences is an appropriate approach.

          • J Mann says:

            Oh, I totally agree that radical consequentialism sometimes seems to lead to intiuitively horrifying results. (Of course, I’m not a consequentialist, in part because of that, so my understanding is definitely imperfect.)

            One remaining quibble: I don’t think the alternative to slavery is tribal populations – I’d say it was something more like the agriculture and commerce that we saw in the free states.

    • EchoChaos says:

      It seems like any chain of moral reasoning that can be applied to slavery and return a “morally okay” must be flawed.

      Why?

      Unless one of your moral axioms is “slavery is never okay”, there are surely offsetting conditions that would make slavery okay in some situations.

      You don’t even have to change your opinion on historical slavery for this to be true.

      • Nicholas says:

        I am referring specifically to the actual historical industry of chattel slavery as practiced in the untied states in the 17 and 18 hundreds, not every possible hypothetical form and conception of slavery. I didn’t think I needed specify, but I realize now that was a mistake on my part.

        Whatever your moral heuristics, if they don’t point strongly to to slavery (actual, historical slavery) is wrong, I think you’ve made some critical errors.

        • EchoChaos says:

          I am referring specifically to the actual historical industry of chattel slavery as practiced in the untied states in the 17 and 18 hundreds, not every possible hypothetical form and conception of slavery.

          I also agree that American chattel slavery was immoral, and strongly retroactively support efforts to peacefully eradicate it like the American Colonization Society.

          Whatever your moral heuristics, if they don’t point strongly to to slavery (actual, historical slavery) is wrong, I think you’ve made some critical errors.

          Actual historical slavery covers a very wide swath, from basically indentured servitude of limited time period (Biblical slavery) to long-term indentured servitude with technical training and possibility of release (Roman slavery) to racially determined chattel slavery with the possibility of release (American slavery) to dehumanizing mutilation and oppression (Islamic slavery).

          It’s perfectly reasonable to have an ethical system that draws a line at a point encompassing some of those and not others. For me, Roman slavery is probably about net neutral, especially in a society without a social safety net, lots of people who would’ve starved ended up as slaves instead. Biblical slavery is a positive good.

          • koreindian says:

            The American Colonization Society was an institution that sent freedmen to Africa, and did not have much at all to do with the “peaceful eradication” of slavery.

            What did you mean here?

            Edit: I see you’ve been temp banned for advocating white nationalism, and am now not confused about why you framed your point this way.

  19. shlemaz says:

    Two well-studied areas where humans are placed in extremely poor conditions are slavery and terminal disease.

    There can’t possibly be a better analogue than the concentration camps, can there? Weird to me that this was not the example he went with as neither slavery nor terminal disease seem in any way comparable.

    • aristides says:

      Do we have good data on concentrate camp suicide rates? I already do not trust the reported suicide rate of slaves (though to be honest I could give a just so story for it being higher or lower than reported), how can I trust Nazi Germany’s statistics? You could go with other concentration camps, like the Japanese internment camps under FDR, but I doubt that was significant. Actually a quick google search has a Wash Post articleclaiming suicide rates quadrupled. I’m on the meat eating side, but I admit this fact gives me pause. I will seriously consider cutting factory farmed chicken from my diet. Also, FDR is even worse of a human being than I thought.

  20. EchoChaos says:

    Thanks for this. It was well written and researched, and despite the fact that I am a meat eater, likely to remain a meat eater and disagree with how to prioritize non-human life.

    It did in fact make me update my priors in terms of meat that I will purchase, though. I don’t eat pork, but I do eat chicken, and this made me substantially less likely to eat chicken and more likely to eat beef at the margins.

  21. J Mann says:

    I enjoyed this one too – thanks for doing it.

    IMHO, these two show the possibilities and limits of the ACC model, which is close to the Wikipedia model and has some similar results.

    The essays do a very nice job of identifying some facts that some people with differing viewpoints can agree on – others can then argue those aren’t the most relevant facts to the question or that it’s possible to disagree with them, but I still feel like I’m learning a lot from each essay.

    Identifying the authors’ initial and final viewpoints is very helpful and interesting.

  22. PedroS says:

    I enjoyed the obvious care taken and the way the authors tried to quantify it all with a model. IMO light-years above yesterday’s entry and the best ACC ever in SSC. I do have one criticism : statistics regarding the prevalence of the specific factory farming practices criticized herein were conspicuously missing, as well as any information regarding their use in non-US markets.

  23. newcom says:

    I’m impressed, this seems a really nice way of quantifying the pro’s and cons of meat eating.
    While the tool you created is probably the most useful part of this ACC, I also enjoyed the part about the ethics of animal suffering.
    Strong entry IMO

  24. J.D. Sockinger says:

    I’ve been vegan for about 15 years, so I was really looking forward to reading this adversarial collaboration. But I found it disappointing (disclaimer: I skimmed it without reading it carefully).

    The authors seem to limit their focus on meat-eating (or vegetarianism), rather than the consumption of all animal products. But as philosopher/lawyer Gary Francione has long argued (persuasively, in my opinion), there is no morally relevant distinction between eating flesh vs. eating dairy or eggs. Both of the latter industries also involve the killing of animals. Dairy cows are slaughtered after they become unsuitable for producing milk. Male chicks are either suffocated or ground-up alive. And egg-laying hens are also slaughtered after their productivity wanes.

    Also, the focus on suffering is misguided. The fundamental issue isn’t how we treat animals, but whether we can use them for our own purposes. I believe it’s morally wrong for us to exploit sentient creatures, even if we treat them humanely in the process. (But if we’re going to talk about suffering, we should acknowledge that there is unspeakable cruelty involved in the production of both milk and eggs.)

    So, all in all, this essay didn’t shed much light on the issue of animal exploitation.

    • aristides says:

      There is a danger in broadening a topic too much. It is easier to persuade someone to take baby steps than large ones. I’ve eaten broiler chickens my entire life, despite reading many others articles trying to convince me it was wrong. I’m Now planning on going home with to my wife and Discussing what steps we can take to eliminate Factory Farmed meat. If this article was half on Veganism, it would persuade me much less. I have chickens in my backyard I raise for eggs, and though you might consider that I am exploiting the chickens, i doubt you would convince me.

      You have identified that there is room for an adversarial collaboration between a vegetarian and a vegan. They would agree on a lot of premises, so it’d be an interesting read. Similarly, I hoped they would discuss non-factory farmed chicken, but I accept that would have broadened the topic too much, especially since that is just 1% of the chicken population. It’s best to have a narrow topic if possible.

      • J.D. Sockinger says:

        It is easier to persuade someone to take baby steps than large ones

        This “baby steps” argument comes up a lot in discussions of animal advocacy, but I don’t find it persuasive. It’s like telling domestic abusers that they should reduce the frequency of beatings and use smaller sticks. If we’re going to take animal interests seriously (and I think we should), then we should advocate for veganism as a moral baseline. Otherwise, you end up with a situation in which animal agriculture becomes even more entrenched, because people think that they’ve solved the problem.

        • Thegnskald says:

          I read this.

          I wouldn’t read the collaboration you wanted.

          • onyomi says:

            It’s interesting; if I recall correctly what seemed like a problem with the ACs last time was that they felt a bit insufficiently collaborative: two sides disagree on an issue, both present a case… and the conclusion is there’s some good arguments on both sides.

            This time people are complaining that (the first two, at least), are insufficiently adversarial–that it seems like one side wasn’t represented or that it was a collaboration between one person who felt strongly steamrolling another person who felt m’eh.

            While I still think the circumcision one gave ethical, cultual, and psychological considerations short shrift, I was impressed they sort of managed to come up with an answer, as did this one.

            I won’t say I wouldn’t read it, but I would not expect to find a collaboration between e.g. J.D. Sockinger (who seems to be a militant vegan based on his posts in this thread) and HarmlessFrog (who seems to be a militant carnivore based on his posts in this thread) to be as helpful as this collaboration–they’re just too far apart to begin with (though, I have to admit the idea of a Dr. Greger-Dr. Baker debate does quite appeal to me; there is a Dr. McDougall v. Dr. Atkins telephone “debate” out there, but I can’t strongly recommend it as shedding much light).

            Which is not to say I think sides with really strong disagreements shouldn’t try to collaborate or hear each other out; I’m just saying there’s a balance to be struck between “adversarial” and “collaboration” and getting it right seems hard. This entry I think did almost as good a job as could be expected.

          • onyomi says:

            Added to note:

            The above is not in any way intended to impugn J. D. Sockinger or HarmlessFrog’s ability to cooperate or reason or write a good piece; I would also be interested to see an AC from collaborators beginning further apart like that. I just expect it would be very hard to fulfill the “collaborative” part of the definition as opposed to just saying “here’s a bunch of arguments from one side and here’s a bunch from the other.”

      • J Mann says:

        +1. I’d say sometimes it’s productive to argue using the existing ethical framework and priors of your audience.

        I already oppose cruelty to animals, but not limits on their liberty. It’s a much easier road to convince me that factory farming (a) involves unnecessary cruelty and (b) is at least health neutral and probably health negative for my family that it is to convince me to change my ethical framework to place a higher value on animal liberty.

        (And you can always try that argument next ACC – I’d like to read it!)

    • Solra Bizna says:

      But as philosopher/lawyer Gary Francione has long argued (persuasively, in my opinion), there is no morally relevant distinction between eating flesh vs. eating dairy or eggs. … Dairy cows are slaughtered after they become unsuitable for producing milk. Male chicks are either suffocated or ground-up alive. And egg-laying hens are also slaughtered after their productivity wanes.

      I’ve known several people who kept pet hens. As long as the hens kept laying, they kept eating the eggs. Eventually, the hens stopped laying and “reverted” to just being unproductive pets until their natural deaths. The moral distinction between this and factory farming—for eggs or meat—is obvious, at least to me.

      Lumping together all possible forms of exploitation of sentient creatures as morally indistinguishable means judging those petkeepers as morally equivalent to factory farmers. This is dangerous black-and-white thinking.

      (Yes, I realize that for every happy pet hen there was a rooster who was ground-up or suffocated or what-have-you.)

    • sty_silver says:

      Also, the focus on suffering is misguided. The fundamental issue isn’t how we treat animals, but whether we can use them for our own purposes.

      According to your moral theory. I, on the other hand, would be completely uninterested in the results of applying any non-utilitarian principle.

      I guess you know this, I’m just a bit annoyed at the phrasing. I think it’s pretty likely that the authors are aware that there are other ethical theories out there and chose not to include them, don’t you think?

    • Froolow says:

      (I’m the vegetarian collaborator)

      I agree completely with your criticism of the collaboration – it would be more complete if we talked about the exploitation of animals in its totality. This might include eating them, eating products derived from them, using them for medical research, using them for cosmetic research, keeping them in zoos, keeping them as pets and so on, so in that hypothetical essay a harms perspective probably would have been incomplete. However my opinion was that the case for vegetarianism:

      1) Was strong and simple, and therefore it might be possible to convince some people to change their lifestyles

      2) Represented the most harmful example of animal exploitation (in its prevalence, if not its cruelty)

      Introducing other topics would also have turned an already-long piece into something almost unreadable in its broadness.

      I obviously haven’t succeeded in convincing everyone in the comments that eating meat is net harmful, and therefore I am glad I didn’t try to take on a more extreme argument like ‘eating any animal products is harmful’ – I think if I lacked the skill to make the case compellingly for vegetarianism there is no way I would have been able to make it compellingly for veganism

    • Nicholas says:

      I haven’t read any vegan-Marxist philosophy, so I’m curious about why you find it moral to exploit non sentient life, and if those conditions could be effectively replicated or transferred to sentient forms.

  25. meltedcheesefondue says:

    Thanks! This was fascinating and instructive, especially about consciousness.

    One quibble:

    >We instead take the consequentialist view that there is a symmetric value in actualizing the existence of conscious creatures that want to be alive (again, a farmed animal’s practical alternative is non-existence).

    There’s no reason that the consequentialist view has to be symmetric in existence/non-existence. Consider for example this consequentialist utility: https://www.lesswrong.com/posts/Ee29dFnPhaeRmYdMy/example-population-ethics-ordered-discounted-utility (there are others with similar results).

    More simply, you can take total utilitarianism and add a negative term for each death. You might think that’s not a worthwhile thing to do, but it is entirely consequentialist.

  26. DragonMilk says:

    Anyone else’s takeaway that of surprise that the average American only eats 7% of a cow and 29% of a pig per year?

    It was my impression that my wife and I consume multiple cows annually. If anything, this makes me feel *better* about meat-eating…

    • aristides says:

      Doublecheck that you are an average American. I calculated my numbers One year, and I was triple the chickens and a fraction of cows and pigs. Cows is also an odd one to calculate, since certain cuts of meat are more popular than others, many of which are left for the dogs. Eating more expensive, and therefore more popular cuts of meat probably has a larger impact on the cattle industry than buying the cheap cuts that are harder to find buyers for. It’s weird on the margins.

      • DragonMilk says:

        I do wonder what the calculation is – if average pounds of beef consumed divided by average weight of a cow, then it’s definitely understating.

        If the average american’s consumption of beef represents only 7% of the meat on a cow per year, then…cows must be way bigger than I imagined

        • Thegnskald says:

          A little over a thousand pounds at slaughter, on average.

          • DragonMilk says:

            But how many pounds of that is edible?

            That figure suggests to me that the 7% is definitely on the entire cow. 70 pounds is about 1.35 pounds per week, which is a sensible amount of consumption.

          • Thegnskald says:

            “Edible” may not be the right metric.

            Everything the cow gets used for is trading off against other things. I have no idea what they’d use in dog food if the non-human-edible portions of cows weren’t used, to pick one example, but it might be the low-value portions of chickens that are currently shipped to China for human consumption.

            So the non-human-edible bits are probably contributing to human food supply indirectly.

        • Froolow says:

          The calculation is average weight of [meat] consumed divided by ‘carcass mass’ of [animal that produces that meat]

          ‘Carcass mass’ is a technical term with a specific definition for each animal, but from what I can tell attempts to get at the amount of animal a human might pay to eat – so excluding blood, bones and meat that is not economically viable to recover, but not making any other adjustments for eg the quality of different cuts of meat

          Speaking approximately, 50-60% of the weight of an animal at slaughter will be ‘carcass mass’

  27. ButYouDisagree says:

    In the discussion on which animals are conscious, I’m surprised that they didn’t reference the Cambridge Declaration On Consciousness, which I understand to be an oft-cited “consensus of the academic field” statement, or Luke Muehlhauser’s Report on Consciousness and Moral Patienthood, which is a similar project to the authors’ section on consciousness, but much, much more thorough.
    Both of these works seem to ascribe consciousness more broadly than the authors do. The Cambridge Declaration says:

    Nonhuman animals, including all mammals and birds, and many other creatures, including octopuses, also possess [conscious] neurological substrates

    (bolding added)

    Muelhauser concludes with a “Probability of consciousness of a sort I intuitively morally care about” of 80% for chickens (same as for cows), 70% for rainbow trout, and 25% for fruit flies.

    • ACC2019 Omnivore says:

      These were covered in preliminary research, and we referenced the Cambridge declaration in an earlier draft but that got cut to make space and focus on the relevant areas of uncertainty: fish and chicken, and so we only linked to the single best source we could find on each of these.

  28. This really misses the actual reasons why the anti-meat argument is wrong, perhaps there was no one arguing that side who held a strong position and had actually thought about it.

    People in the west don’t eat horses, dogs, and cats, and there’s a good reason for that. We cooperate with those animals, and eating them would count as “defecting” in that cooperation. We eat certain other animals because we are purely using them, without cooperating with them. Thus the whole question is framed wrong: the attempt to calculate “how much does it hurt for them” is from the standpoint of cooperation: you are talking as though you are cooperating with them. In other words it is advocating being an always-cooperate bot, and that’s a bad idea morally, and it’s bad for you personally. That question is irrelevant: the question is overall would WE be better off if we cooperated with those animals the way we cooperate with horses or dogs.

    • thevoiceofthevoid says:

      Many people, including myself, care about the well-being of agents that we do not expect to ever provide us any benefit. When I have disposable income, I plan to donate significant portions of it to e.g. orgs that purchase anti-malaria bed nets. I don’t expect to have any significant risk of contracting malaria; nor do I expect any significant personal benefit from the people whose lives are saved by bed nets. Yet, I still plan to donate, because I care about the well-being of other conscious beings as a terminal goal. In short, I am not a rational egoist.

      From a moral standpoint where you value the suffering or happiness of others as an objective good, it makes perfect sense to worry about the well-being of farm animals, even if you could never “cooperate” with them in the same way we do with horses or dogs.

      Disclaimer: Though I suspect that eating factory-farmed meat is probably somewhere between morally questionable and morally abhorrent, I still regularly eat meat. Articles like this bring the cognitive dissonance into focus.

      • Evan Þ says:

        From a moral standpoint where you value the suffering or happiness of others as an objective good, it makes perfect sense to worry about the well-being of farm animals

        From that standpoint, I agree. However, have you considered a moral philosophy that values the wellbeing of humans as an objective good, but not that of nonhuman animals?

        • Dacyn says:

          Well then the question is, why humans? Is it because of some features that they have? Many such features are also present in animals, to a lesser extent.

          If it’s not because of anything specific then it seems like just… species discrimination.

        • thevoiceofthevoid says:

          I have considered such a philosophy. However, I find it lacking for a couple reasons. Instinctively, I find the torture of cats and dogs repugnant, and am disgusted by the pictures of factory farms. However, I have similar instinctive reactions toward images of lifesaving surgeries which I unquestionably morally endorse, so this isn’t the best metric. Philosophically, if a non-human animal has a conscious experience and feels happiness and suffering in a similar way to humans, I see no more reason to exclude it from my moral consideration than I see to exclude humans of a different race, or those who speak a different language, or those with intellectual disabilities. Whether they do experience human-like consciousness is a matter of empirical debate.

  29. Moorlock says:

    [I]f animals suffer under current farming standards to the point of preferring non-existence then there is a moral burden on meat eaters to justify eating them.

    This seems to be the bedrock foundation you have chosen for evaluating the moral side of your argument, and a moment’s thought shows it’s repugnant. Would you argue that way about, say, human slavery? That people who had been bred to be enslaved would present a moral burden only if it could be shown that they would prefer non-existence to slavery? Would you argue that way if you were one of the factory-farmed animals in question (“boy, if this ever gets so bad that I’d prefer non-existence then those who are subjecting me to these conditions will have a real moral burden on their hands”)?

    The thought experiments in which naive utilitarians are asked to imagine an evil genie who creates zillions of horribly suffering unfortunates who nonetheless value living over dying and then asks to be congratulated for its philanthropy — these are meant as reductios, not blueprints.

    • Said Achmiz says:

      Would you argue that way if you were one of the factory-farmed animals in question (“boy, if this ever gets so bad that I’d prefer non-existence then those who are subjecting me to these conditions will have a real moral burden on their hands”)?

      No, I wouldn’t.

      … because, in that scenario, I’d be a cow or a chicken or something, and thus incapable of arguing anything or holding any view. You know, what with being a dumb animal and all.

      (How do people overlook this? It’s a pretty important aspect of the whole situation, wouldn’t you say?)

    • blacktrance says:

      I think the condition is chosen for its strength (i.e. “if animal lives are that bad, then meat-eaters definitely have a moral burden…”). It’s a sufficient condition, not a necessary one, and whether their existence is preferable to death provides a target for analysis.

  30. jcrox says:

    Does anyone have some good sources on the effects of pornography consumption, for the relative layman? Anywhere someone’s explained the literature particularly clearly?

  31. thisheavenlyconjugation says:

    I don’t think this is successful in the sense of reaching a conclusive conclusion or being likely to persuade anyone (except people who weren’t aware of the relative differences in quality of life of different animals). But given the magnitude of the question I think it’s a strong attempt. IMO the authors would’ve done better to look at a narrower part of the question in more depth.

    • Thegnskald says:

      Useful conclusions out of it:

      Beef is probably the most ethical meat of the three discussed, maybe even being morally neutral to consume, from an animal perspective.

      That’s more of a conclusion than I would have expected, and I’m impressed, particularly since I think they did a reasonable job taking into account a wide variety of potential perspectives.

      • caryatis says:

        Do the authors agree with this potential conclusion?

        • Thegnskald says:

          They assigned negative values to days of chicken and pig lives, and a small positive value to days of cow lives, so if they don’t, I’d need an explanation.

        • ACC2019 Omnivore says:

          Beef is probably the most ethical meat of the three discussed, maybe even being morally neutral to consume, from an animal perspective.

          Yes, that is the conclusion there was least disagreement on/highest confidence in.

          • eccdogg says:

            And it also seemed that you both agreed that eating most seafood was no problem. Usually not factory farmed and except perhaps in the case of cephalopods unlikely to be conscious.

            Would that be a correct interpretation?

          • ACC2019 Omnivore says:

            @eccdogg Yeah we agreed eating fish is fine as well.

        • Froolow says:

          (I’m the vegetarian collaborator)

          After completing the collaboration, I agree unreservedly that beef as it is currently produced is the most ethical of the three being discussed – but I am less confident than my partner that eating beef is morally neutral / beneficial (mostly because I am less persuaded by the habituation literature than they are)

          Having said that, it was extremely surprising to me that beef came out so much better than chicken and pork – this was the issue on which I changed my mind the most

    • aristides says:

      It’s useful to get less biased numbers. I’ve read arguments on both sides, and came down on eating meat is fine. Reading this knowing a meat eater certified these numbers is making me strongly consider eliminating factory farmed chicken from my diet. I’ve read about cruel conditions of chickens before, but I didn’t trust them, since they value chickens lives much more than I do. This shows that if you give them a moral weight any higher than 0, you should probably stop eating factory farmed chicken.

  32. Dave Baker says:

    As much as I love adversarial collaboration, and appreciate the evident work the authors put into this…

    I find this article profoundly disappointing, because it ignores the single best argument that eating meat does not produce net harm: Namely, that the nature of the supply chain in large developed countries is such that an individual’s consumption habits have literally no impact on the amount of meat that is factory farmed.

    See the work on this by the excellent Mark Budolfson, e.g.:

    But now consider this question: if you purchase animal products at a supermarket or restaurant, are the welfare effects more like those of buying a steak at Jimmy’s or more like those of acquiring a steak through dumpster diving? Conventional wisdom among consequentialist moral philosophers says that the effects are more like eating at Jimmy’s; however, the empirical facts suggest that they maybe more like dumpster diving, because it is virtually impossible for an individual’s consumption of animal products at supermarkets and restaurants to have any effect on the number of animals that suffer and the extent of that suffering, just as it is virtually impossible for an individual’s consumption of products acquired through dumpster diving to have any effect on animal welfare.

    (“Is it Wrong to Eat Meat from Factory Farms? If so, Why?”)

    • Nietzsche says:

      Yes. Very much this. The same is true of environmental impact. It is really disturbing how little people understand collective action problems and persist in the fairy-tale belief that their individual action has any impact whatsoever. It is true that if I slaughter Bessy the cow and eat her, then to whatever extent that is a morally-assessable action, it is on me. However, global-scale environmental impacts are not on me in any way. If a vegetarian wants to persuade me that I personally should become a vegetarian, telling me what would happen if all 7.5 billion people were vegetarians has no probative force at all.

      • I generally think the green people shouldn’t be listened to, but the same logic could be used to morally excuse Holocaust perpetrators.

        • Nietzsche says:

          That is incorrect. A Holocaust perpetrator is morally responsible for their individual actions. They are not responsible for the entire Holocaust.

      • Solra Bizna says:

        I turn off lights when I’m not using them, configure my computers for low power usage, set the thermostat high in the summer, et cetera. This saves me a few hundred dollars of energy a year, while costing me relatively little in time, effort, and inconvenience. I have also been known to pick up trash from sidewalks, parking lots, and streets, and carry it a considerable distance to a trash receptacle—an action with absolutely no economic benefit to myself.

        I don’t do things like these because I have some belief that my individual actions make a dent in energy supply or the environment. I do them because I would rather live in a world where more people do these things. I am the only person whose actions I can directly control, and doing these things myself is therefore the most direct way I have to increase the number of people in the world who do them.

        Moreover, a few other people have observed my lifestyle and modified theirs in a positive direction based on my example; in a world where nobody changes their behavior because everyone knows their individual choices don’t matter, this doesn’t happen and nothing ever gets better.

        • Nietzsche says:

          This seems inconsistent to me. You write “I don’t do things like these because I have some belief that my individual actions make a dent” but then go on to say how you think they make a dent: other people follow your example, e.g. Of course, you and a handful of followers have no impact on global-level problems. You write, “I would rather live in a world where more people do these things”, but I don’t think that’s actually true. You want to live in a world where EVERYONE does these things. Sure, me too. But my singular action won’t bring that world about. There’s no reason to care that more people act as you do. There is only a reason to care that a very, very large number of people do.

          • Solra Bizna says:

            You write, “I would rather live in a world where more people do these things”, but I don’t think that’s actually true.

            Actually, it is. I’d rather live in a world where 0.0001% of people don’t eat babies¹ than a world where 0% of people don’t eat babies. And I’d rather live in a world where 0.001% of people don’t eat babies than a world where 0.0001% of people don’t eat babies.

            In all such worlds, me choosing not to eat babies doesn’t impact the global rate of baby consumption in a detectable way, and even if two or three other people decide to stop eating babies after seeing me not eat babies that still hasn’t affected the global rate of baby consumption in a detectable way. But a world where everyone says “Everyone else is eating babies, I might as well eat babies” is a world where everyone will eat babies forever, whereas a world where any people say “Everyone else is eating babies, but I think I’d rather not eat babies” is a world where any people at all don’t eat babies. That makes a future where most people don’t eat babies at least possible. No individual non-baby-eater’s choice will “bring that world about”, true, but that’s not the point; individual people making non-baby-eating choices is still a strictly necessary component of that world coming about.

            1. Baby eating not intended as a metaphor for meat consumption, in spite of the article under which this is posted.

        • JohnBuridan says:

          Be the change you want to see in the world!

          I think the proper distinction is between moral obligation and moral freedom or praiseworthiness. This is a distinction our community often forgets to make. One judge that it is good to take a course of action, without that course of action being obligatory.

          It can be praiseworthy to cut back meat consumption and pick up trash, even when there is not a moral obligation to do so.

    • thisheavenlyconjugation says:

      He mentions “empirical facts” but doesn’t seem to give any, rather there just seems to be an assumption that this market is inefficient. I seem to remember reading that actual empirical facts suggest the opposite — production of meat is actually very sensitive to changes in demand — but I can’t find the source (think it was on GWWC).

      • Dave Baker says:

        Would be interested to see a link on this. I think his thought is that it’s highly implausible that the market is so sensitive to demand that one individual’s consumption would be noticed. It’s possible for a market to be very efficient without being that efficient.

    • When ranchers who own their own grazing land decide
      how many cattle to raise, their decisions are sensitive to their own
      financial situation, the number of cattle their land can support, the
      expected price of any additional feed that will be needed, bull semen
      and other “raw materials” that go into cattle production, and the
      expected price that the cattle will fetch when they are ultimately sold to
      feedlots. Of these, small changes in the last item—the price that cattle
      will fetch at the feedlot—are of the least importance, because insofar as
      ranchers judge that capital should be invested in raising cattle rather
      than other investments, they will tend to raise as many cattle as they
      can afford to breed and feed within that budget, letting the ultimate
      extent of their profits fall where it may at the feedlot.

      This argument is meaningless and tautological.

    • RC-cola-and-a-moon-pie says:

      I’m going to read this. But pending that, does this argument leave open the rejoinder that one is nonetheless morally obligated to work toward the legal prohibition of meat consumption (or certain types of meat consumption, like meat from factory farms or whatever)? Even if an individual’s consumption patterns are irrelevant, surely a whole nation’s would matter.

      • Dave Baker says:

        Yes, it does. The only question addressed is whether it’s wrong to eat meat under current conditions. I would support a ban myself, although banning meat farming makes more sense than banning meat consumption.

    • skybrian says:

      There is a metaphor I like to use here: if a volcano has erupted and lava is flowing towards the village, there is no point in getting a shovel and trying to build a dike; the first step is to warn the village. It may be unclear what to do, but it’s clear you can’t do it alone.

      So, I agree on the importance of collective and relative insignificance of individual action. But I don’t think that changes the argument much? You still have to decide whether you’re in favor of working to reduce meat consumption. Collectively, the supply chain wouldn’t exist without consumer demand.

      It means we should place more weight on advocacy and systematic changes and somewhat less on individualistic effort. But there is still sometimes value in setting a good example, for collective-action reasons.

    • Anatid says:

      From the link

      waste, inefficiency, and other forms of slack may seem to ensure that the real expected marginal effect of an individual’s consumption is essentially zero, because the change in the signal received at the production end of the supply based on a change in a single individual’s consumption decisions is almost certainly zero.

      This doesn’t seem like it could be right. If say 20% of people stopped buying chickens, presumably 20% fewer chickens would get raised. If we imagine each of those people stopping one at a time, it would be impossible to resolve the effect of each individual person stopping because each person only represents like 0.0000001% of the demand, but *on average* each 0.0000001% reduction in demand must result in 0.0000001% fewer chickens being raised. It might well be true that because of “slack” the system only changes in larger discrete steps rather than 0.0000001% at a time, like maybe the system only reacts at all after 100000 people stop and then it readjusts by a full 0.01% . But even in that model each person has a small chance to cause a large discrete step, and *on average* each person has an effect.

      Maybe there is an issue here with the meaning of “expected effect”. In statistics, it means “mean effect”. Maybe Budolfson is instead using it to mean “most common effect”. But I think the mean effect is the relevant thing for this moral decision, not the most common effect.

      • Grantford says:

        This exactly. I don’t see how the argument that Dave mentions can really hold any water. Maybe any individual’s economic decision will go unnoticed, but the decisions of a million consumers will be noticed. By not consuming meat, you may have no impact, but you may instead be the millionth consumer to make such a choice and finally push demand over the edge such that the market notices that a million people aren’t demanding meat.

        • RC-cola-and-a-moon-pie says:

          If the objection is that you may be a one-in-a-million tipping point for some marginal production calculation then I think you would have to discount the expected value of the harm from eating meat to reflect the one in a million chance that you will cause it to occur, and even then only count the presumably very small marginal effect that that would have on meat production. I think you’d end up with a fairly tiny number, which could easily be offset by the expected benefits of meat consumption.

          • Grantford says:

            I don’t know the actual number of people required to go vegetarian before the market notices and makes a discrete reduction in its supply in response to the lower demand. I suspect it’s less than a million in real life; let’s just call that number X. What I’m saying is that there is a 1-1/X chance that your choice will have no market impact. But there is a 1/X chance that your choice will have the huge impact of reducing the number of animals produced by however many animals X people would eat. This averages out to the intuitive expectation that your choice to not eat meat reduces the number of animals produced by exactly the number of animals that one person eats.
            If an average meat eater eats a couple dozen chickens annually, and if chickens’ lives are as bad as they seem, this difference does not seem tiny.
            I feel like Anatid might have explained it better than I did, but I think we’re saying the same thing. So if my explanation doesn’t make sense, maybe theirs will.

          • Dacyn says:

            No, the point is that the effect in the one-in-a-million case is not small because it has to be a million times as large as it would be in the case where you are sure to have an effect. So the two “million”s cancel.

      • sty_silver says:

        I was scrolling down this thread desperately hoping someone would have made this point. Thank you. It doesn’t actually make any difference how quickly or reliably the market reacts; the expected effect of changing consumption is always identical.

        Or, I mean, maybe there does exist some mechanism that would change this – but the burden of proof is very much on the side claiming that there is an argument here.

      • thevoiceofthevoid says:

        +1 to you and Grantford, exactly what I was gonna say. With the basic economic assumptions that the amount of meat produced by the ag industry is about equal to the amount demanded by consumers, and at least depends on it somehow (albeit with various amounts of slack and thresholds at various points)…the expected effect of buying/not buying one unit of meat should be one more/less unit of meat produced on average. Even if the most likely effect of a lifetime of vegetarianism is nothing, as he claims, the average effect should be a lifetime’s worth less meat produced, unless the giant profit-focused agricultural companies are somehow dumb enough to produce meat with a complete disregard for the consumer demand for meat.

        • Said Achmiz says:

          Even if the most likely effect of a lifetime of vegetarianism is nothing, as he claims, the average effect should be a lifetime’s worth less meat produced, unless the giant profit-focused agricultural companies are somehow dumb enough to produce meat with a complete disregard for the consumer demand for meat.

          This would seem to require a low (but not too low) probability outcome where a lifetime of vegetarianism results in much less than a lifetime’s worth less meat produced. I don’t think the math works out.

          • thevoiceofthevoid says:

            The basic argument goes like this:
            Assume you buy your chickens directly from the grocery store, which orders them in batches of 100 from the slaughterhouse every week. The grocery store ideally wants to buy exactly as many chickens as it sells; if it buys too few, it will run out; if it buys too many, some will spoil on the shelves. However, exact week-to-week demand varies, and they can only purchase in multiples of 100. So, they might follow an algorithm of: Purchase however many chickens were sold last week, rounded up to the next multiple of 100. Say you previously bought a chicken a week, but are now considering the impact of not buying it anymore. On any week, your decision is very unlikely to impact their purchase from the slaughterhouse. They’ll buy 600 chickens next week whether they sold 572 or 573, so buying chicken #573 or leaving it on the shelf makes no difference. But, on one out of every hundred shopping trips, they’re right on the margin, and your decision to not buy chicken #701 means that they’ll only buy 700 instead of 800 chickens, saving 100 chickens from the slaughter. The factor of 100 cancels out in the expected value: You have a 1% chance of saving 100 chickens, giving an average of 1 chicken/week saved by forgoing the purchase of 1 chicken/week.

            Now, there are more levels than that in the supply chain of meat production in the real world, and more “slack” in each level, so the actual chance of an effect on the number of animals raised and slaughtered is much, much smaller than 1% for a single purchase. Budolfson argues that despite what an economically-inclined ethicist might think, ranchers will be completely insensitive to any individual consumer decision, continuing to produce the same number of cattle even with moderate decreases in price of cattle at the feedlot. (That might make sense if the marginal cost curve is discontinuous, @DavidFriedman please correct me if I’m wrong: graph) According to Budolfson, since any individual’s decisions won’t outweigh the noise in that system, even a lifetime of vegetarianism might never cause any rancher to decide to raise fewer cattle. However, I contest that if individual ranches don’t change production in response to market changes in price, we must then account for the tiny probability that your lifetime of not buying beef reduces the demand just enough to drop the marginal revenue to the point where a rancher goes out of business entirely (or decides not to buy additional land to raise thousands more cattle per year). According to the macro supply and demand curves, we should expect that to happen with a probability that exactly cancels out: a 1/[large number] probability of your lifetime of vegetarianism reducing the demand for cattle enough for a ranch producing [large number] times the number of cattle you’d have eaten in your lifetime to go out of business or fail to open. Expected value ~= number of cattle you’d have eaten. If not, then the giant profit-focused agricultural companies are not responding to market incentives, which seems unlikely.

      • Blueberry pie says:

        I broadly agree, but I think a big thing you need to account for is elasticity – me not buying meat (which I try to reduce) means the meat gets cheaper and might mean some other people who would not buy meat before will buy it. Since there is a large population of people in poorer countries that would eat more meat if they had the money, I don’t think you can ignore this.

        I still believe not eating meat personally is a good thing. And it can have other effects than changing the demand for meat by one person’s consumption.

        For example, it can make vegatarianism high-status (partially already happening), so people will try to emulate it. In my country, people not eating meat used to be mocked – and they still are – but much less so.

        A small number of vegetarians can also help develop business models and infrastructure around plant-based food – for example, in my country it is slowly becoming more common for restaurants to have sensible, well prepared and tasty vegetarian options, which was rare some 20 years ago. Vegetarian-only restaurants used to be unthinkable but now exist in small numbers. So it is easier to not eat meat today than it was before, because of small number of people turning vegetarian individually. Will this have large effects on society? Hard to say, but it is at least plausible.

      • Dave Baker says:

        I think what’s missing from this calculus are two other pieces of knowledge you possess when making your decision: (1) There is currently enough “slack” in the supply chain that we are not currently anywhere near a “threshold.” (2) The probability of 100,000 other people joining you in spontaneously becoming vegetarians so as to bring us up to the next threshold is approximately zero.

        Compare the following sort of case: You are one of 100,000 people who stand surrounding a glass tank with a man inside. Each of you has a glass of water and the Great Leader is exhorting you all to pour your water into the tank, collectively filling it and drowning the man. You know everyone but you is brainwashed and the chance of any of them disobeying is zero. Sure, you can decide not to pour your water, but that doesn’t mean the expected utility of doing so is 1/100,000 of the expected utility if everyone declines to pour. It is zero.

        Of course the probability of an average person in your country spontaneously becoming vegetarian is not literally zero. But it is much closer to zero than the probability of you becoming vegetarian (if you’re reading this). So you need to take that into account.

        (I think Budolfson could have made it clearer that his argument is depending on assumption (2).)

        • thisheavenlyconjugation says:

          But the 100,000 people don’t have to become vegetarian simultaneously. If you’re considering the case where you are the only person ever to become vegetarian, then Budolfson would be right. However, in the real world plenty of people will have converted before you (and you can expect plenty to do so afterwards).

        • sty_silver says:

          Say there are 10n meat eaters. Now n of them stop eating meat. The meat production will have to decrease by a tenth. Let’s call that total amount X.

          Now say we order the n people who stopped eating meat by (p_1, …, p_n). Now we can assign each person an impact: how much did they reduce meat production?

          Since the market is inelastic, the number is going to be zero for the vast majority of the p_i. Probably for over 99.9% of the p_i. But not for all of them, since the total impact is X. That X must be distributed somehow over those n people. So almost all of them will have 0, but then some will have a lot more than X/n, which is the mean impact. Maybe the first 7523 didn’t do anything, but then number 7524 reduced the total production by roughly 7524*(X/n).

          Now suppose some guy stops eating meat. They don’t know where in this sequence they are. They have absolutely no clue. Every position is equally likely. So most likely they have 0 impact. But there’s also a chance that they are one of the ones that tip the scale and reduce the meat production by a whole lot. Such people do exist; as shown above, the market has to react at some point, and that point has to be after one of the people who stopped.

          So if it takes 7524 people to cause a change, then roughly that person has 7523/7524 chance of doing 0 and 1/7524 chance of reducing the total meat production by 7524*(X/n), which is 7524 as high as the mean impact. This ends up with an expected impact of exactly X/n.

          Obviously, this works regardless of which numbers you plug in. The expected impact of stopping to eat meat is always exactly as large as it would be if the market immediately adjusted to every customer.

          So inelasticity is not even a weak argument or a partially valid argument. It’s not an argument at all. For your personal expected impact, it’s completely and utterly irrelevant.

          Edit: In case you’re tempted to reply by pointing out that there are no n people who will stop to eat meat – that doesn’t change the calculation. Even if the meat production only ever goes up, you still have a very small chance of having prevented one of those times that it went up, and the numbers come out exactly the same.

  33. A Lorenzen says:

    I really liked the part on consciousness and I think it could convert me from vegetarianism to pescetarianism. Still, I would have liked a more throughrough exploration of the conditions under which holding animals is ethical.

    The central question is whether factory farmed animal lives are worth living; the realistic alternative to meat eating is not a better life but for those animals to not exist in the first place.

    This seems to imply that we are acting ethically if and only if the animals prefer living. I totally buy the ‘only if’ part, e.g. “Animals would rather die -> we should feel bad”. But the ‘if’ part (“Animals don’t hate this -> we are fine”) feels very weird.

    Assume we were talking about humans and you could grow some in a lab. Is it okay to grow and hold these as slaves (which is better than non-living apparently)? I don’t think so. Assume we were talking about sweet little dog puppies that you found outside and saved from the cold. Is it okay to abuse them once they are big enough to attack the postman? The legal code in my area disagrees.

    In the rest of the analysis, they only use the ‘only if’ part, but it makes this analysis unnecessarily vulnerable to people claiming that non-factory-farming-meat is fine.

  34. caryatis says:

    Most of this was wasted on me because I don’t think animals’ interests matter. It sounds like neither of the collaborators had this belief. Questions about the health of meat-eating are more complicated than is acknowledged here. I was not convinced this collaboration fully addressed arguments that eating meat is affirmatively good for adult humans.

    • Thegnskald says:

      The environmentalism and health sections discuss everything I can see that it would be meaningful to talk about from those two perspectives, including preferences for eating meat.

      • caryatis says:

        They tried, but it was too broad of a topic. And they didn’t address arguments that eating meat is affirmatively good for health, as opposed to « I enjoy it. »

        • thevoiceofthevoid says:

          Agreed. They made a nod towards “meat has nutrients plants lack”, acknowledged the inherent flaws in all observational studies touting purported benefits of vegetarianism, and….drew their conclusion from a single study touting purported benefits of vegetarianism.

    • RC-cola-and-a-moon-pie says:

      There have been several comments like this, and they surprise me. I do think there’s an element of truth here in that any analysis which fails to differentiate among species is doomed to error. But do you really not think that animals’ interests matter? Are you ok with the torture of a cow (or a puppy) for fun? Repeal of all laws against animal cruelty? No problem with the worst forms of old-fashioned vivisection? Hard to believe that anyone would wholly write off the suffering of obviously sentient, complex organisms like that.

      • Cameron Harwick gives the correct answer here: https://cameronharwick.com/blog/why-i-am-not-a-vegetarian/

        The authors of the article take for granted that you are an always-cooperate bot, and that’s bad.

        • RC-cola-and-a-moon-pie says:

          Thanks. I appreciate the link but I really don’t find this persuasive. Adopting a cooperation-based grounding for moral consideration seems just as arbitrary as the article claims the sentience-based ground is, and it has the disadvantage of producing wildly counter-intuitive results that few would find satisfactory (like there being no reason to care about the suffering of a cow that is being tortured for fun, etc.) In fairness, I take an entirely different view of ethics but putting myself into the shoes of this author I don’t really understand the motivation for this approach apart from a genetic fallacy that the author professes to reject.

        • thevoiceofthevoid says:

          I think that article starts with an implicit assumption of egoism: it presumes that the only purpose of moral behavior is to improve one’s own well-being in the long term via mutually-beneficial cooperation. I disagree.

          As Eliezer poetically explains, our goals and moral values are not the same as the “goals” of the process of natural selection that created us. And one of our “godshatter” traits, which admittedly would make Evolution shudder in horror if it weren’t an unconscious abstract process, is our capacity for compassion towards people who can’t possibly reciprocate the good we do towards them. I care about other people as a terminal value, and if you tell me that helping people who can’t help me back (i.e. altruism) is against my rational self-interest, you’re entirely missing the point. Likewise, if I’m convinced that an animal is conscious and experiences happiness and suffering, I will care about its well-being. If you argue that cows don’t cooperate with us, you don’t understand what values I’m trying to optimize.

          To quote my friend’s laptop sticker: “I don’t know how to explain to you that you should care about other people.” It’s a terminal value for me, not just an instrumental one.

          • caryatis says:

            >Likewise, if I’m convinced that an animal is conscious and experiences happiness and suffering, I will care about its well-being.

            So which of those things matters to you? Because it seems to me that consciousness, experiencing happiness, and experiencing suffering are different things, and I’m not sure that an animal that can do one of those things could do the others.

          • thevoiceofthevoid says:

            @caryatis

            A better phrasing would have been “if an animal is sentient in a similar way to humans and thus has experiences that are analogous to our own conscious happiness and suffering”. I don’t see how you could “experience” anything without being conscious. The main thing that matters to me is “happiness or suffering analogous to our own”, and I think some kind of consciousness/sentience is clearly a prerequisite for that. That is, a fly might exhibit operant conditioning to positive or negative stimuli which you might call “pleasure” or “pain”, but without a conscious experience of such I’d be reluctant to call it “happiness” or “suffering”.

          • thevoiceofthevoid says:

            @caryatis

            I replied at much greater length to your comment below.

          • gkai says:

            Having consciousness as a central and mandatory feature to judge suffering is a slippery slope in itself.
            You have a lot of nasty deep questions in there, first one is what is consciousness? Think Turing test, where only the actions are used because examining internal state is something we can not do even for other humans (in fact, it’s even debatable what really happen when you examine your own thinking process, it seems neuroscience now think of consciousness as a after-the-fact rationalization instead of the older and more popular root decision making process.
            Other food for though are the classical moral questions: what can you do to someone who is not conscious, or someone who is conscious but will not remember anything. Some drugs makes those questions quite practical, regarding anesthesia for example….

            So saying an animal really suffer only of it is conscious seems a very shaky position. In a way, it’s more honest to admit arbitrary position like I will not stand a dog or my my pet fish suffering, but lobsters I do not care, they are far too tasty.

          • thevoiceofthevoid says:

            @gkai

            Having consciousness as a central and mandatory feature to judge suffering is a slippery slope in itself. You have a lot of nasty deep questions in there, first one is what is consciousness?

            I define my terms explicitly and make my argument in more depth below, but briefly: Consciousness is having internal experiences; it’s what you have when you’re awake but not when you’re (dreamlessly) asleep. Sentience is the capacity for consciousness. Though I use the terms somewhat interchangeably, if I’m being strict you’re sometimes conscious but always sentient (assuming you’re an adult human who sleeps).

            Think Turing test, where only the actions are used because examining internal state is something we can not do even for other humans (in fact, it’s even debatable what really happen when you examine your own thinking process, it seems neuroscience now think of consciousness as a after-the-fact rationalization instead of the older and more popular root decision making process.

            For other humans, it’s actually quite easy: ask them if they’re having internal experiences! While it’s not conclusive proof, I’m generally inclined to take them at their word, since they look like the same kind of creature as me and I know I have an internal experience of what it’s like to be me. They also exhibit plenty of other behaviors consistent with consciousness, such as discussing at length the intricacies of their emotions.

            For animals, it’s a harder problem, since they can’t talk. However, as this article discusses, we can still appeal to neurophysiological similarity to known conscious beings (ourselves) and, where that fails, observed behaviors that are consistent with conscious experiences and would be difficult to explain in a non-sentient creature. In any case, just because something is difficult to define or measure doesn’t mean it can’t be morally relevant! GiveWell is an organization entirely devoted to trying to measure difficult-to-define-and-measure things that are morally relevant, like “How many lives will be saved on average if I donate to charity X?”

            Other food for though are the classical moral questions: what can you do to someone who is not conscious, or someone who is conscious but will not remember anything. Some drugs makes those questions quite practical, regarding anesthesia for example….

            My pro-happiness and anti-suffering stances are not my only moral values! I have an immense weight in my moral calculus for any sentient being’s desire not to have horrible things done to them while unconscious.

            So saying an animal really suffer only of it is conscious seems a very shaky position. In a way, it’s more honest to admit arbitrary position like I will not stand a dog or my my pet fish suffering, but lobsters I do not care, they are far too tasty.

            As I define suffering (“experiencing pain as a negative sensation”), you can’t suffer without being conscious by definition. Maybe I’m defining it more broadly then you do, but if sentience is “having experiences” then it’s obviously a prerequisite for “having unpleasant experiences”.

          • gkai says:

            I think your position is fair enough, so I will try to explain how my opinion differ:

            We agree that consciousness is difficult to define, and you partly solve this by using max consciousness in your life time…Maybe max consciousness you should extend that max value for all member of the species, this will get rid of a nasty trap having some humans (heavy mental handicap) less conscious and thus less empathy-worthy…

            This is convenient for humans, but do not help much for animals. If consciousness is basically inner experience and is difficult to get from observing action, I maintain we have a problem. Animal Intelligence assessment try to focus on actions observed during concrete problem solving, and there we get some birds and some cephalopods doing better than some mammals despite brain organisation difference (sometime massive difference). I thus seems difficult to use brain organisation as a proxy for consciousness…
            Frankly, apart from humans who have the benefit of langage and of being the same species as the observer, I do not see how to assess consciousness separately from intelligence, the attempts I have seen to differentiate focus on the mirror test (or variants), with usually do not differ from conclusions of general problem solving, and in the case it do, it may simply be because evolution have favored visual recognition of individuals in the species (instead of olfactive, auditory, or no need for recognition of individuals)…
            So my position is that consciousness as a concept separated from intelligence is doubtful, and moreover I do not see how it is specially linked to suffering evaluation.
            I feel it’s more a fudge factor that allows to valuate more animals that humans naturally feel more empathic to, in the case natural empathy do not align with tested intelligence….

            Regarding your final point, I have trouble differentiating what you call experience from simple memory, and most animals can form memories that will influence their future behavior, even insects…I do not know if there is a fundamental difference between automatons with memories and conscious actors, but I tend to think at one point they become indiscernable…
            Probably general AI will be needed to really close this old philosophical debate

          • I’m not going to give a full response to this because I’m not in the mood, but equating cooperation and selfishness is one of the more weird claims I’ve seen from people.

      • caryatis says:

        Yes, I mean what I say. Animals’ moral claims are so far below those of humans as to be nonexistent for all practical purposes.

        I wouldn’t want to torture an animal (or see one tortured) because it would make me feel bad. Not because of the anima’s rights. As for whether that should be illegal, I would think about the benefits and costs of such laws for humans. Animal experimentation seems clearly good (some vivisection was needlessly cruel, some made valuable contributions to science). I would be concerned that allowing animal torture with absolutely no restrictions would be bad for humans, in the same way that murderers have been observed to start with animal cruelty. Not sure about that, but it’s something to consider.

        • RC-cola-and-a-moon-pie says:

          Thanks for the reply. Candidly, wholly disregarding the suffering of a violently tortured animal (i.e., caring only about it insofar as it impacts people) seems extremely callous to me. But I hear what you’re saying!

          • caryatis says:

            If you think animals have no rights, then squeamishness about their suffering has no moral value.

          • dionisos says:

            If you think animals have no rights, then squeamishness about their suffering has no moral value.

            This doesn’t follow except if your moral system is entirely based on rights.
            Personally I don’t care about rights at all.

        • ana53294 says:

          Do you care about chimps?

          Some researchers claim some primates have low-human levels of IQ (70 or so for some of the highest).

          • caryatis says:

            If the average chimp had a 70 IQ, I would think chimps deserved some rights. Much below that, probably not…but any IQ number I could come up with would be an arbitrary line-drawing exercise.

        • thevoiceofthevoid says:

          Do you think that animals are likely not sentient? Or do you disagree that we ought to treat non-human sentient creatures similarly to humans?

          • caryatis says:

            I guess my other response to you gets at this. Despite having read more than one book by Singer, I don’t know what “sentience” means. It seems to lump together capacity to feel pain or pleasure, capacity to feel emotions, and consciousness, which are different things.

            I do think (many) animals feel pain, but it seems arbitrary to say that ability to feel pain grants a being rights.

          • thevoiceofthevoid says:

            @caryatis
            Allow me to use this thread for a second attempt to define my terms more clearly:

            Pain: Reaction to a negative stimulus [1]
            Pleasure: Reaction to a positive stimulus
            Conscious: Having an internal experience [2]
            Sentient: Has the capacity for consciousness [3]
            Suffering: Conscious experience of pain
            Happiness: Conscious experience of pleasure

            Now that I’ve got all my terms defined, let me think of how I’d best define what gives an entity moral weight in my book.

            [1] It look like the technical term would more properly be “nociception“, with “pain” often being defined as I’ve defined “suffering” (i.e. with reference to the conscious experience). However, since I’d never heard of the technical term before and I can’t find an equivalent one for reaction to positive stimuli, I’ll stick with “pain” and “pleasure” for the non-necessarily-conscious reactions.

            [2] This definition of “conscious” is kind of pushing the buck to “experience”, so my best extensive definition of consciousness would be “The thing that you have when you’re awake, and kind of have when you’re dreaming, but don’t have when you’re in dreamless sleep or knocked out.”

            [3] I use “sentient” and “conscious” somewhat interchangeably, but to be rigorous: Sentient beings are those that are conscious sometimes. They can be temporarily unconscious, e.g. when in dreamless sleep, but going to sleep doesn’t make you temporarily non-sentient.

          • thevoiceofthevoid says:

            As I said above, I think happiness is a moral good in and of itself and suffering is a moral bad in and of itself. All else equal, I’d like to see a world with more happiness and less suffering, though I have other moral values that sometimes trump them–I’m against wireheading, for example, since that seems likely to lead to the loss of creativity, love, and all the other things I value. I only care about pain and pleasure (as defined above) when they occur in creatures that can experience them as happiness or suffering (i.e. sentient creatures). If a plant sends out a pheromone in response to its leaf being plucked but has no internal experience of the sensation, I couldn’t care less about it.

            I think it’s usually morally good to cause an existing being to feel happy, and almost always morally wrong to cause it to suffer. I’m much less sure about the ethics of bringing new beings into existence. Creating a creature with a happier-than-average life is probably good; creating a creature with a life that is worse than average but better than death I’m not sure about; but creating a creature with a life worse than death, that suffers more than it’s happy, has got to be morally abhorrent. This ACC agrees with me on the last point, and takes the stance that creating creatures with lives barely better than nonexistence is morally good, which I’m not sure whether I agree with.

            Of course, that discussion is all moot if the animals in question aren’t sentient and thus don’t experience their pain or pleasure as conscious suffering or happiness. And determining whether animals are sentient is difficult, since they can’t talk to tell us whether they’re having internal experiences. I think that while emotionless sentient agents are theoretically possible, it’s incredibly unlikely that a biological organism would evolve to be sentient without experiencing happiness or suffering. This is because the obvious evolutionary purpose of consciousness as far as I can tell is enabling intelligent reasoning and learning in pursuit of basic evolutionary drives, with the drives themselves encoded as suffering and happiness (i.e. punishment and reward signals). However, a basic organism could easily have reflex responses to pain without being sentient.

            The authors of this ACC address this question at length, and I find myself generally agreeing with their analysis. I know some other people are conscious because they tell me so; I know people with locked-in syndrome are conscious because they still display the same brain activity as verbally-confirmed conscious people. (I generalize this result to people who I’ve never talked to nor seen their EEG, for obvious inductive reasoning.) Other great apes have very similar brain structures to us, display complex behaviors, and have even picked up rudimentary sign language, so I don’t think it’s much of a stretch to guess that they’re also sentient. I think there’s a strong case for mammals being conscious; as the AAC’ers mention, all mammals have a cerebral cortex, which seems to be the “conscious part” of our brains, and many display various complex behaviors that are difficult to explain as simple reflex responses to stimuli. Perhaps there may even be a continuum of sentience/consciousness, with different species (and maybe humans at different stages of development) at different points along it. I’m not sure to what extent a lower moral weighting for beings lower on the spectrum would be acceptable. We see children differently from adults in terms of moral values like autonomy, but we obviously aren’t ok with torturing babies before they’re old enough to form memories.

  35. algorizmi says:

    Many good points already raised by other commenters such as the difference between grazing and arable land, intelligence of fish, false dichotomy, and weakness of health evidence.

    Have yet to see mentioned that: Beef is much more resource intensive than chicken; and fishing, either wild-caught or most forms of aquaculture, have major ecological impact. While battery chickens may be some of the worst from an animal welfare perspective, they are among the best in an efficiency/sustainability context.

    Also because of the difference between ranching and CAFOs, 90th-percentile for CO2 emission is not the group as 90% for land use, I’d expect them to be anti-correlated if anything.

  36. eric23 says:

    Generally well done, thorough and balanced.

    The Adventist study you link to appears convincing at first glance, but seems to be lacking in a couple ways. It doesn’t appear to distinguish between people who eat beef (which many studies suggest is harmful to health) and people who eat other forms of meat (where studies do not consistently say that). It doesn’t account for the possibility that people joined vegetarian churches in part because they were more willing to accept vegetarianism to begin with (relative to those who joined similar but non-vegetarian churches).

    A lot of meat farming uses land which could not be productively used for crops (e.g. too arid or mountainous)

    It appears that the calculation is very different for different demographics. Perhaps children should best have meat to avoid possible malnutrition, while middle aged fat people should avoid meat because it contributes significantly to their weight gain.

    I read that debeaking is only done for egg-laying chickens, not eating-chickens? In general, I hear the conditions for these two types of chickens in factory-farms are quite different (though egg-chickens probably have it better), and should not be confused.

    • Froolow says:

      (I’m the vegetarian collaborator)

      I think we both agreed the Adventist studies were the best of a very bad and confusing bunch of weak studies on nutrition, and at least somewhat consistent with the majority of other (even worse) studies – I don’t think either of us would want to defend them further than that

      On the issue of debeaking, I think you are probably right and the collaboration is probably wrong. There seems to be significant confusion about whether and when broiler chickens are debeaked – but in general upon revisiting the claim after reading your comment it seems that the sources claiming that broiler chickens ARE debeaked are mostly biased towards making factory farming sound as bad as possible, and the sources claiming that they ARE NOT tend to follow up the claim by arguing that broiler chickens are usually killed before they grow a big enough beak for it to matter, which sounds plausible. So I can’t absolutely confirm that you are correct, but I think upon checking the claim we made is too strong

      But you are absolutely right that egg and meat chickens have very different experiences – they are frequently confused and after doing this collaboration I wouldn’t trust any source that isn’t explicit about which type they are talking about

  37. skybrian says:

    This is a good effort, but I feel like another article could be written devoted to whether fish are conscious, better than the one they linked to. I sort of agree that they probably aren’t, but I’m not confident in this and it’s pretty close to the heart of the argument.

    • Grantford says:

      I also felt that the conclusion of fish not being conscious was arrived at a bit too soon. They did link to an paper entirely about consciousness in fish, so it may be that that paper was very convincing to the ACC authors.

  38. Chad_Nine says:

    You were willing to link to disturbing images of factory farm conditions, but not to a circumcision being performed?

    • Purplehermann says:

      This makes sense to me

    • Dacyn says:

      I imagine it was less relevant to the argument, since the immediate physical suffering of children as they are being circumcised wasn’t addressed. Also, it doesn’t make so much sense to say “you” when you are referring to two separate groups of people.

    • eric23 says:

      Pictures of children’s genitals are rather frowned upon in our society

  39. Loris says:

    Comparing across countries is difficult, but it seems that America is slightly more industrialized than the EU. My best estimate is that the difference is not significant enough to make a moral difference.

    The first sentence may be technically true, but at least here in the UK the USA is notorious for its very poor food animal welfare legislation and intensive rearing of animals. Example article. Conditions for such animals in the EU may not be as good as some (including myself) would want, but they’re probably at least significantly better than those in the USA.

  40. graehl says:

    The idea that which Adventist church you attend is confound-free doesn’t fly at all with me.

  41. graehl says:

    If there’s pasture grass existing for free, sheep or cows seem a no-brainer.

  42. Purplehermann says:

    In the future, maybe collaborators should put out a outline of the points they plan to cover a week or so after they start, so the peanut gallery can throw in their two cents?
    The crowd sourcing would help make less vectors be missed by the pair.
    [Edit: this is not an attempt to throw shade. A lot of commenters seem to believe (in both colabs so far) that the colabs left what to be desired, and have specific ideas about what else should be addressed, so I think collabs can gain from a group brainstorm early on]

    • Thegnskald says:

      Eh. I’ll give this one an enthusiastic thumbs up. The circumcision article, a thumbs down.

      My biases going into both articles was pro-meat-eating and Anti-circumcision.

      I feel this article was fairly representative of the possible arguments that could meaningfully be had, nevermind the peanut gallery. It is meaningful that there are complaints from both sides about this article, but only from one side on the other.

      You can’t please everybody, but if the people upset are representative of both sides of an issue, you’ve either done something right, or absolutely horribly wrong, and it is pretty clear they didn’t do anything horribly wrong.

      • Purplehermann says:

        Reading the comments here, both sides seem to believe there were additional (meaningful) points to be discussed.
        This would be enough to draw fire from both sides, even without doing a bad job.

        At the end of the day the writers will decide what they want to address and how, but the peanut gallery here is a resource that could improve the end result.

  43. Armadillo Daffodil says:

    Thanks for a comprehensive and well-written entry. Well done!

    Now, one thesis I think could be examined in more detail is this:

    The central question is whether factory farmed animal lives are worth living; the realistic alternative to meat eating is not a better life but for those animals to not exist in the first place.

    This brings to mind the mere addition paradox, and what Parfit referred to as “the repugnant conclusion”. Specifically, if we consider a life worth living to have positive value, and keep adding up the values, it seems that a world with a sufficiently large number of individuals whose lives are barely worth living could be “better” than a world with far fewer individuals whose lives were quite good.

    My ethical intuition says that this can’t be the case; it’s wrong to bring something into existence if you expect their life to be “bad”, where “bad” is some value a fair bit greater than “barely worth living”.

    • Grantford says:

      Hm. If the difference between a bad life and a good one lies at some value above “barely worth living”, aren’t you just re-locating the zero point of the utility spectrum? And if you did that, then wouldn’t a point just above this new zero point also be “barely worth living”?

      I like your introduction of Parfit-style population ethics into the discussion, but I’m a bit unclear what you have in mind here. Unless all you are saying is that marginally good lives aren’t actually worth living.

    • Anatid says:

      Or similarly: if bringing cows into existence is morally good because their lives are worth living, then we ought to at least consider whether there is some cheap and environmentally friendly way to raise a bunch more happy cows, not to eat but just because it would be moral to create those lives.

      Unless I missed it, neither the authors nor any of the commenters have suggested doing this, which maybe suggests that no one seriously believes in the exact moral system that the authors assume in their essay?

  44. Grantford says:

    Thank you to the authors for clearly putting a lot of work into this ACC! I mostly thought that they did a good, thorough job, but I do have a few critiques on some matters that seemed underdeveloped.

    The authors’ conclusions on consciousness seem a bit premature. The identity of the neural correlates of consciousness is a hugely controversial matter among neuroscientists, so concluding that fish are almost certainly not conscious after only a few paragraphs of discussion seems a bit hasty. However, I should admit that I have not yet read the linked article on phenomenal consciousness in fish, and that article may be doing a lot of work for the ACC authors.

    The ACC authors seem to make a couple of underlying assumptions about which type of utilitarianism to apply. The first of these is the assumption of preference utilitarianism over hedonistic utilitarianism. Their analysis seems to accept that if an animal lives a bad life but does not actively wish for that life to end, then the animal’s preference takes precedence. However, as the authors mention, it seems plausible that evolution has ‘tricked’ animals into being death-averse even in cases where their lives are objectively bad. I am a bit doubtful that we should privilege this evolutionarily derived preference over the actual goodness or badness of the experiences in an animal’s life.

    The second assumption about which type of utilitarianism to apply concerns a distinction whose name I do not know but which is hinted at by Parfit’s Mere Addition Paradox, mentioned above by user Armadillo Daffodil. I will refer to the two positions as ‘natalist’ and ‘non-natalist’ utilitarianism since I do not know whether they have generally accepted labels already. Natalist utilitarianism assumes that the total utility in the universe is what matters, and if you can bring new beings into existence that will live good lives, then you absolutely should, even if this leads to Parfit’s Repugnant Conclusion. Non-natalist utilitarianism assumes that only the utility or preferences of beings that have existed, do exist, or will exist matters. If you can avoid the Repugnant Conclusion by not adding any new beings and focusing on improving the lives of beings that already exist, then this is what you should do. That the meat-eating author considered buying meat and throwing it away just to encourage the creation of more beings suggests an underlying assumption of natalist utilitarianism, but the ACC didn’t argue why this background position would be preferable to non-natalist utilitarianism. If cows, pigs, and sheep actually do live net positive lives, natalist utilitarianism would encourage consumption of these animals, while non-natalist utilitarianism would find no value in that.

  45. deciusbrutus says:

    So, this ‘adversarial’ collaboration was between two people who didn’t have significant disagreement about the value of the conscious experience of chickens? Or at least, they were able to agree on values for it, as opposed to ranges.

    • Froolow says:

      (I’m the vegetarian collaborator)

      We disagree hugely about the value to assign to the conscious experience of chickens! However, we wanted to set our base case up to represent ‘typical’ beliefs about animal consciousness, and I conceded that vegetarians were not typical with respect to their beliefs about animals (in the sense that there are many more meat eaters than vegetarians – a meat eater is the ‘median eater’ in that sense). Therefore, the number we settled on was that proposed by the omnivore, and I agreed that that seemed like a true approximation of the beliefs of a typical person

      • Grantford says:

        I think that it was a reasonable decision to take the typical (meat-eating) person’s perceptive for the sake of charity.

        However, it should also be noted that a meat-eater’s beliefs about animal consciousness are likely much less reliable than a vegetarian’s beliefs in this area due to wishful thinking. The meat-eater is personally incentivized to believe that animals are not conscious (or only have low-value conscious experiences) so that they can justify continuing to enjoy their meat consumption.

        • TripleS says:

          It should also be noted that vegetarians also face a variety of personal and social incentives to choose a moral framework that most closely aligns with their preferences – saving the lives of entities that are in fact just fluffy people from a qualia perspective is way more satisfying than saving something with only a small fraction of a human’s consciousness (and way easier to use as an argument to convert others). Plus, since we can see that there are a broad variety of schools of vegetarianism and veganism with different taboos followed, we can conclude that even if wishful thinking is the philosophical equivalent of vitamin B12, most of them are *still* wrong about some or all of the values they’re using. Just look at Hinduism, a religion with millions of vegetarians or vegans who will tell you that actually a cow is the next step up the cosmic ladder (something most westerners of any palate would find unlikely and unhelpful in this debate), and it becomes apparent that unless we’re presupposing that the only way to eat meat is to lie to yourself and find empty ways to justify your atrocities, assuming one side has more reasons or avenues to be irrational than the other is the biggest wish of all.

          • Grantford says:

            Actually, vegetarians do suffer from the same motivated reasoning as meat-eaters: (most) vegetarians also want to eat meat. For vegetarians, though, this bias towards supposing that animals aren’t conscious isn’t enough to overcome whatever high (relative to meat eaters) priors they have on animal consciousness. Because of the desire to eat meat and to be in an epistemic position where they are at least more likely to switch to meat-eating in the future, it seems plausible that vegetarians also tend to underestimate animal consciousness. They just do this less than meat eaters do.

            The most reliable beliefs on animal consciousness would probably come from people who naturally dislike the taste of meat and other animal products and thus have no desire to consume those products in the first place. Such people, having no dog in the fight, wouldn’t suffer from this specific instance of wishful thinking.

          • thevoiceofthevoid says:

            Plus, since we can see that there are a broad variety of schools of vegetarianism and veganism with different taboos followed, we can conclude that even if wishful thinking is the philosophical equivalent of vitamin B12, most of them are *still* wrong about some or all of the values they’re using.

            I think this is largely explained by the fact that there are a variety of different reasons that people go veg*n, and not all of them have to do with reducing animal suffering. Religious prohibitions optimize for appeasing whatever deity’s in you x-thousand-year-old book of choice, and shouldn’t necessarily be expected to track with any of these rational reasons to cut out meat. Semi-vegetarians who care primarily about the environment or their personal health should probably substitute chicken for red meat; those who care primarily about animal suffering (and think a chicken is anywhere near as conscious as a cow) should do the opposite. Both are rational for optimizing the values they care about. Vegetarians who still eat eggs or dairy might be blissfully unaware that egg hens are in conditions similarly terrible to broiler hens, or might be deontologically opposed to killing animals but ok with raising them for products (and/or blissfully unaware of the amount of killing involved in milk/egg production), or have simply made the tradeoff that milk isn’t thaaaat bad. Or again, they might be vegetarian for health or environmental reasons and thus facing an entirely different set of moral incentives and tradeoffs. And if all you’re saying is “vegetarians don’t all have the exact same moral values!” then…well obviously people don’t all have the exact same moral values, no matter what subgroup of humanity you choose.

      • caryatis says:

        I would have liked to see a conversation between you and someone who thought animals’ conscious experience has no value.

    • eric23 says:

      I’m getting the feeling that a set of arguments and rebuttals, produced by thoughtful people, would work better than an adversarial collaboration on a single document.

  46. Lorec says:

    (Disclaimer: I only read the ethics sections because of my prior that ethics overwhelms environment and health.)

    Thanks to the authors for writing about animal farming ethics, which to a utilitarian-adjacent who thinks farmed animals suffer a lot is the most neglected issue in discourse. Your emphasis on the suffering-intensive lives of broiler chickens is appropriate and far too uncommon.

    My surface concerns, which others have also addressed:

    1) If your support of factory farming hinges on whether animals’ happiness set point is positive/”worth living,” and you apply this principle to your entire moral framework, that implies the repugnant conclusion that the world could be improved by breeding to stretch resources until each person’s standard of living was at the minimum positive value. I personally find that situation unattractive, and also want to avoid speciesism and status quo bias (where the status quo is factory farming) in my moral reasoning, but this is an entirely normative question and people have all sorts of moral intuitions regarding animals so the most I can do is point out soft “contradictions.” Personally, I don’t have a comprehensive ethical position, but given my belief in significant animal consciousness, consistency requires me to treat factory farming the same qualitative way I would treat a horribly oppressive, torturous and murderous human social institution.

    2) You lean heavily on neuroanatomical analogies in evaluating degree of consciousnesses. You justify this using the differing impact of human brain parts on conscious thought. I know next to nothing about neuroanatomy but for now I’ll bluntly state my strong intuition that the degree to which another agent’s consciousness resembles mine can be inferred from the similarity of our input-output patterns. This slots consciousness neatly into a world where 1. every phenotype as complex as consciousness serves an essential purpose, and 2. other organisms appear to vary independently of exact hardware plan in how similar to mine are the purposes their brains must serve. As far as I am aware the issue of fish consciousness is contested and I am not sure that any evidence weak enough to leave brain scientists divided will overcome this extremely convenient prior.

    I may do a more thorough reading and review of your research and write a full post response somewhere.

  47. Etoile says:

    A piece of the ethics of meat from the meat side: meat was once upon a time somewhat of a luxury for the average family: the availability of abundant and affordable meat to so many people can be argued as a massive improvement in quality of life, and therefore should be weighed against the animal suffering – especially since there is a case to be made that you cannot supply the world with affordable meat without resorting to some degree of factory farming.

    • thevoiceofthevoid says:

      I think that’s essentially what’s considered in the “psychological costs” section–how much do you enjoy eating meat, quantified by how much you’d have to be paid to give it up.

  48. Subb4k says:

    I think that’s a solid work. I share the concerns of some other commenters on your implicit natalist perspective (and the associated repugnant conclusion), but I believe this can be more or less adjusted away in the parameters. You could, for example, set all positive experiences for animals to 0.

    As a pescetarian who relatively recently switched from a “I eat everything except mammals” diet, I have a few suggestions on how to improve the decision aid should you decide to do so:
    * Your environmental impact calculator currently does not depend on the amount of meat one currently consumes. I guess the argument is that what matters is what humanity does, and therefore one should look at the average contribution from meat-eaters rather than what one can gain by going vegetarian. However, it is not explained how this average contribution is obtained (using an average US diet? an average Western diet?). Furthermore, I believe knowing how much of those environmental resources are used for you specifically matters for ethical decision, especially if your maxim is to act in such a way that you are morally better than average”.
    * It would be relevant to include milk and eggs in it, especially since it sounds like egg-laying hens have even worse lives so you can’t just convert the number of eggs you eat to a number of hens needed to lay those eggs over their lives (subtracting the meat from their eventual slaughter if you don’t eat it).
    * Similarly, it would be relevant to include chickens raised outside. I don’t know if it’s a thing in the US, but here in France it’s very easy in any supermarket (and most corner stores) to buy chicken/eggs that are labeled either “open air” or “free range”, which indicates they get to live at least half of their lifetime outside the shed during daytime (the difference being whether the chickens have a fence-in area with minimal 2m² per chicken, or no fence at all). I don’t know what are the conditions for the 3/4 of their lives spent inside a shed (nights during the first half, all of it during the second half), but I assume the calculations would work out more similar to cows than to shed chickens.

  49. Chalid says:

    Since praise is generally undersupplied in comment sections, I just want to say that I thought this was really great work. Thank you very much to David G and Froolow for all their effort.

  50. Acleveralias says:

    I was a lifelong omnivore until I tried switching to a vegan diet about two years ago. I tried this diet for purely health / weight loss reasons and didn’t find the ethical case particularly compelling. After (mostly) eating vegan for about a year, I noticed that my moral intuitions around eating meat had shifted quite a bit. When presented with meat (especially steak), I was more aware that I was looking at a chunk of an animal that had previously been alive. Probably still tastes good, but also kinda gross. It also started to seem that causing animals to suffer/die so that I could eat them when I could rather easily just choose to eat something else was a kinda shitty thing to be doing. Being the sort of person that doesn’t frivolously inflict harm feels like clearly a more “right” way to be. The ethical case against eating meat had gone from ‘easily dismissed, eye roll’ to ‘obviously correct’.

    The thing to notice here is that my ethical perspective didn’t change due to any new evidence or arguments being presented to me. The thing that changed is that I was no longer creating rationalizations to defend my behavior. I had already changed my behavior for unrelated reasons, so I was no longer clinging to arguments because they told me my actions were 100% correct and I didn’t need to change anything. In retrospect, the bias is pretty obvious. I’m probably still biased now (aren’t we all?) but I think there’s a lesson here in being aware of your own motivated reasoning.

    • thevoiceofthevoid says:

      I’m pretty sure that if/when lab-grown meat tastes as good and is a cheap as real meat, I’ll join the rest of society in looking back in horror at the time when we kept animals as conscious as our dogs in cages too small for them to turn around so we could raise, slaughter, and eat them on an industrial scale. But for now…changing my diet is hard, I’m not totally convinced that pigs are morally relevant, and there’s enough distance between me and the factory farms to just not think about it most of the time.

      • Acleveralias says:

        I found cutting meat out of my diet to be much easier to do than I expected it to be, though it probably depends a lot of the sort of food you like eating now. In a lot of cases it is as easy as selecting “tofu” instead of “chicken” in your curry or choosing “veggie burger” instead of “burger”. I’ve been really impressed by how good/widespread the fake meat products have gotten in the last few years.

        If you’re at all interested in seeing how easy/hard it would be for you, see if you can cut down your meat intake by 50% for a month – could be a good way to discover what the alternatives are like.

        • thevoiceofthevoid says:

          Chicken would probably be easier. The flavor of most good chicken dishes comes from the sauces and spices, not the chicken itself. But while the impossible burger I tried once was a good approximation of the charred “well-done” burgers I’ll stomach at cookouts, it didn’t come close to the juicy, medium-rare burgers I love. However, as @Chalid correctly points out below, if chickens are conscious then cutting them out will eliminate most of the harm caused by my carnivorous ways. I might try it.

      • Chalid says:

        It is really quite easy to cut chicken out of your diet (or at any rate reduce it to very low levels) and according to this collaboration that alone will do a lot of good.

      • HarmlessFrog says:

        I’m pretty sure that if/when lab-grown meat tastes as good and is a cheap as real meat, I’ll join the rest of society in looking back in horror at the time when we kept animals as conscious as our dogs in cages too small for them to turn around so we could raise, slaughter, and eat them on an industrial scale. But for now…changing my diet is hard, I’m not totally convinced that pigs are morally relevant, and there’s enough distance between me and the factory farms to just not think about it most of the time.

        I don’t trust the lab-grown meat substitutes. It just looks like ultra-processed, nutrient-deficient meal-replacement slop to my heuristics, and I doubt they’ll manage to perfect the process enough to produce something that offers the same nutrition as meat within my lifetime. (Never mind conducting long-term clinical trials to prove its equivalence to real meat.)

        • thevoiceofthevoid says:

          Ultra-processed, yeah sure, but I think the food factories are pretty good at injecting the required nutrients into foods they’re trying to market as “healthy”. See: “Fortified” anything. Milk and cereal don’t naturally have a lot of vitamin D in them but due to factories adding em in they’re probably your main source of it if you live where the sun ain’t all the bright and don’t eat a lot of fatty fish.

    • Skeptical Wolf says:

      One thing I find counter-intuitive is that people seem to find charity harder in discussions about dietary choices and nutrition/health than they do elsewhere. Given the sorry state of nutrition science, I would expect empathy in this field to be easier rather than harder.

      We have pretty good numbers for the total environmental impact of different vehicle choices (certainly much better than we have on nutrition, let alone animal suffering). But when someone chooses to drive a less fuel-efficient vehicle they aren’t usually greeted with the same vehemence or condescension that routinely shows up in any discussion of a weight-loss study.

      Even here, at SSC, what other discussions routinely have “People who disagree with me only do so due to motivated reasoning” and “People on one side of this debate should be assumed to be thinking less clearly about it” show up so commonly?

    • HarmlessFrog says:

      The thing to notice here is that my ethical perspective didn’t change due to any new evidence or arguments being presented to me. The thing that changed is that I was no longer creating rationalizations to defend my behavior. I had already changed my behavior for unrelated reasons, so I was no longer clinging to arguments because they told me my actions were 100% correct and I didn’t need to change anything. In retrospect, the bias is pretty obvious. I’m probably still biased now (aren’t we all?) but I think there’s a lesson here in being aware of your own motivated reasoning.

      How sure are you that you’re not, currently, rationalizing not eating meat, as you were rationalizing eating it before?

    • You’re almost surely on the ball with this one, but just because it’s a personal concern I always have when I read this: Please, please, please make sure you’re supplementing enough vitamin B12. <3

      (Yeah, I’m that person who had a vitamin B12 deficiency despite eating meat, but then made worse by a few times I tried to go vegetarian (no, not even vegan). It was a nightmare that I cannot overstate. Severe depression, lasting cognitive damage (thankfully minor, but noticeable). So don’t be like me, folks!)

      • HarmlessFrog says:

        You’re almost surely on the ball with this one, but just because it’s a personal concern I always have when I read this: Please, please, please make sure you’re supplementing enough vitamin B12. <3

        And test that properly, too. Serum B12 doesn’t work. You need a methylmalonic acid test to determine B12 deficiency.

        https://www.youtube.com/watch?v=9dB7SwbnGbY

  51. janrandom says:

    This is fantastic! It is the first writeup that takes a very comprehensive approach and manages to convert most dimensions into a common unit – money and makes this available as a tool that is easy to adjust and extend. That’s also why I don’t understand the criticism. While the individual results of the authors may be in the middle of the field the framework is powerful enough to calculate everybody’s impact based on the weights.

    I am especially happy to see the area of population dynamics. The one area that I considered important and underexplored and was always looking forward – and here it is.

    > One (very approximate) way is to assume that the human race is polluting the planet as much as it can sustainably (at current technologies it’s a lot worse, but we’re science-optimists), and that the long-term impact of a reduction in pollution is a proportionally inverse change in the equilibrium human population.

    Question to the authors: You don’t give references here and I’m not sure I see what calculations you did here. Do you have more details on the method and numbers?

  52. dionisos says:

    Great Adversarial Collaboration Entries.
    Thanks to the authors.

  53. Cheese says:

    I think this is a good AC. Congratulations to the authors. While it’s not written in an adversarial tone I think that is often the point, a compromise position is arrived on and the bulk is written from that perspective.

    This piece shifted my view to some extent, and will (hopefully) have a few effects including slightly more vegetarian meals, obtaining some chickens for home-based egg production (a viable choice where I live), attempting to increase efforts to find more ethically farmed pork in particular (i’m non-US and I think this is a viable choice) but also feeling slightly less bad about beef and lamb consuption with a similar proviso around attempting to source ethically farmed. I wonder if the environmental aspect of fish consuption is adequately addressed though – overfishing strikes me as a valid concern that i’m not 100% sure is mitigated by farmed alternatives.

    I disagree with a lot of the criticisms in the comments, many of which I think are asking for the inclusion of perspectives from outlying positions, ethical or otherwise, or overwhelming demands for rigour in all areas which would basically turn every AC into a near-fulltime job for the collaborators for a month.

    The only point I agree with would be that some treatment of infant vs child vs adult nutrition might be useful. I think there might be reasonable evidence that animal protein deprivation in infants, and less no children, is highly problematic as compared to adults.

  54. Skeptical Wolf says:

    I was surprised by the claim that vegetarian diets are less expensive. Are there any more details available on where that estimate came from? I have investigated this on multiple occasions and have never been able to put together a vegetarian menu without sacrificing at least one of nutritional completeness, enjoyment of the diners, and cost (cost as a concern aggregates both ingredient cost and cooking time/effort).

    This trend has been softening somewhat in recent years as vegetarian protein products have improved (I quite enjoy impossible burgers, even if I’d never mistake one for beef). However, direct substitute products remain more expensive (boca/morningstar/quorn crumbles are usually about 50% more expensive than 90/10 ground beef where I shop). And non-substitute products require substantially more time, effort, and supplementary ingredients to reach the same level of enjoyability, even before the reduced variety is considered.

    For people who have switched to vegetarianism, what was your experience in this regard? Did you end up spending more or less on food during the time of your transition, and how did this change over a longer time-frame? Did you find your meals to be less enjoyable and/or more difficult to prepare? If so, to what degree was that effect mitigated over time as palette adjustment / habituation / improved cooking skills took effect?

    Thank you to the authors for putting this collaboration together. It did not cause me to make any large behavior updates (for a variety of reasons, most of which do not relate to the quality of the analysis), but I did learn some valuable things by reading it.

    • indigo says:

      I don’t keep a food budget, but I have varied in how much meat I eat over time. It’s always from local farms that treat their animals well, nothing factory-farmed.

      My perception is that it’s pretty simple: When I eat meat, it replaces some of the plants I would be eating in that meal. And that amount of meat is way more expensive.

      I don’t have studies to back that up. It could be that meat somehow replaces a really large amount of vegetables and grains and legumes; perhaps it could be that meat makes me less hungry in some way. But these seem unlikely.

    • onyomi says:

      Vegetarianism is definitely cheaper in absolute terms: you can meet your basic nutritional needs cheaper with a giant sack of rice and a giant sack of beans than you can with any animal product.

      It may not always work out cheaper in practice, especially if you’re trying to appeal to the tastebuds of a food culture with a lot of meat in it, like American or French cuisine. For example, most Americans will find just rice and beans with no meat too bland, so then you end up using some kind of vegetarian sausage meat replacement to fill the perceived void in your beans or chili, and that product is probably more expensive than some cheap pork sausage.

      As home to the largest population of vegetarians Indian food unsurprisingly does vegetarian food as well or better than anyone else. Where you might have missed the sausage and ham hocks in a US-style pot of beans, you don’t miss it so much in a serving of dal, idli and sambar, etc. because the flavor is bulked up with a bunch of spices and other veggies (in addition to ghee and cream at many Indian restaurants–not healthy, most likely, but tasty–most Indian vegetarians do eat dairy).

      • Skeptical Wolf says:

        It may not always work out cheaper in practice, especially if you’re trying to appeal to the tastebuds of a food culture with a lot of meat in it, like American or French cuisine. For example, most Americans will find just rice and beans with no meat too bland, so then you end up using some kind of vegetarian sausage meat replacement to fill the perceived void in your beans or chili, and that product is probably more expensive than some cheap pork sausage.

        This describes my experience exactly. If I was willing to accept “Food just won’t be enjoyable any more” for myself and my household, we could improve a lot on any number of other metrics. Perhaps unfortunately, mealtimes being pleasant is not something I’m willing to give up. So far, the best we’ve been able to do is mixed protein dishes (our primary beans-and-rice entree also uses some sausage).

        In general, I see food decisions as a balance between four primary factors: nutrition, cost, convenience, and taste. It’s usually pretty easy to improve one or two of those by accepting losses in one or two others, but after learning to cook and acquiring a menu broad enough to incorporate most locally-available ingredients, strict improvements have been very difficult to come by. Every equilibrium I’ve found sustainable in that system so far has involved at least some meat and other animal products. Removing those ingredients is introducing an additional optimization criteria, which has pushed the rest of the system into a less desirable equilibrium every time I’ve tried it.

        Eating meat is not a terminal value (at least for me). If I could achieve a better health/cost/convenience/taste result without animal products, I would have become vegan by accident years ago in pursuit of that goal. This is the point I was getting at in my original post: the assumption that the vegetarian diet is cheaper seems to be assuming a large trade-off that is unaccounted for elsewhere.

      • HarmlessFrog says:

        Vegetarianism is definitely cheaper in absolute terms: you can meet your basic nutritional needs cheaper with a giant sack of rice and a giant sack of beans than you can with any animal product.

        If by “basic” you mean “caloric”.

    • Acleveralias says:

      My context: switched to a vegan diet for health reasons, so the switch was towards healthier meals as well as vegan ones.

      Buying meat-substitute ingredients is indeed somewhat more expensive than the real-meat equivalents when preparing meals at home. The vegetarian part of the diet doesn’t really make food prep harder or food significantly less good, but the no-dairy part does (cooking without butter and cheese is somewhat challenging at times).

      Eating out is a different story though – meatless meals tend to be somewhat less expensive on average. The quality and variety of vegetarian options will depend a lot on what kind of restaurant you enjoy:
      *Indian and Thai food is trivially easy to go vegetarian with. Usually 75% or more of the menu is already vegetarian or has a vegetarian option (most curries you can select tofu or vegetables instead of chicken) The flavor of these dishes doesn’t come from the meat, so you pretty much lose nothing when you lose the meat
      *Italian is fairly easy, there’s lots of great veg pasta.
      *Mexican depends on the place. Chipolte and Qdoba both have solid options for a fast lunch.
      *Burgers are super easy – just substitute a veggie patty.
      *BBQ is just the worst, sometimes literally nothing on the menu. These places are optimized for people who don’t care what they eat at all.

      It is a lot easier if you live in a major city, especially one on the west coast. Here in Seattle there are vegan options at most every restaurant. In Florida you’re more likely to find gator on the menu than a veggie burger.

      It certainly got a lot easier over time as I discovered more meat and cheese substitutes, experimenting and seeing what I like and don’t like. I don’t miss beef or chicken at all at this point, but do miss fish sometimes (and will still eat sushi once every few months).

      • Skeptical Wolf says:

        Thank you for sharing your experience. I’m sure region makes a big difference in these options (for reference, I live in central Michigan).

        Have simple substitutions consistently worked out well for you? I find that directly substituting the protein component of almost any dish makes a huge difference (though I don’t eat much curry, so that could be a blind spot in this regard). I have had veggie burgers that I enjoyed, but only when they were prepared in a completely different way than how I normally prepare meat patties; when I tried to just swap in a veggie patty, the bitter notes became the early dominant flavor and the shorter taste duration left an unpleasant bland space in the middle of the profile. On the flip side, the only peanut noodles I’ve really enjoyed were made with tofu, not chicken, so this sort of thing cuts both ways.

        What protein source do you usually use when eating italian? Around here, the main vegetarian option that doesn’t involve a small mountain of cheese is usually eggplant parmesan.

        Historically, BBQ was optimized for taste and cost; it was a way for people to make something enjoyable out of something that would otherwise have been trash. It has obviously developed from there, but the focus on taste remained. It isn’t optimized for people who don’t care what they eat. It’s optimized for people who care very much about what they eat, in a completely different way than you do.

        • Acleveralias says:

          Have simple substitutions consistently worked out well for you?

          Yes, but I’m not typically trying to do anything that elaborate. Tossing veggie-chicken or veggie-burger in a salad or rice bowl works great.

          There’s a whole lot of variety in veggie burgers. Some of them (impossible burger) are attempting to replicate the meat experience as closely as possible, while others (bean burgers) are just an alternative that isn’t intended to be identical. My favorite is Boca, doesn’t taste exactly like a burger, but tastes better than any real meat patty I’ve bought at a store.

          What protein source do you usually use when eating italian?

          There are veggie versions of Italian sausage now, but I’d usually order something based on other flavors (like mushroom). Butternut squash ravioli is incredible, and is something I probably wouldn’t have discovered if I didn’t go vegan.

          In my opinion, people are also way over-concerned with protein consumption. Protein deficiency isn’t really “a thing”, at least not for adults in developed countries. If you aren’t a pro athlete, you don’t need to consume protein-rich food at every meal. I try to optimize for flavor and nutrition-to-calorie ratio.

  55. indigo says:

    I would have liked to see more coverage of factory vs. non-factory farming, and stats on how much the average US person would have to reduce their meat consumption in order to allow for all meat and egg chickens to be raised comfortably.

    “Eating less meat” is a perfectly valid middle ground to discuss.

  56. VirgilKurkjian says:

    Doesn’t the pro-meat argument here prove too much, i.e. doesn’t it suggest that human cannibalism would be acceptable on the grounds that factory farmed humans probably wouldn’t commit suicide most of the time, implicitly suggesting that factory farmed humans would prefer that terrible existence over nonexistence?

    Sorry for horrific thought experiment, honestly not trying to be incendiary. If the authors admit that they would support factory farmed human mean I would be horrified but also would grudgingly admit respect for their intellectual consistency.

    • hnau says:

      With this ethical framework almost any argument can prove too much. For example, according to the analysis, if the same resources that support 1 human in a normal happy life could support 10 happy pigs instead, then we ought to make that substitution and in fact we ought to make it many times over.

    • whereamigoing says:

      My guess is that many positive utilitarians would accept that conclusion, and negative utilitarians would view it as an argument for negative utilitarianism. Personally I think it’s an argument for contractarianism.

      • siwhyatt says:

        It could be argued that most humans do “choose” to live on the farm, albeit as work animals rather then livestock.

        By opting to live in society we exchange our freedom for security. Most people do have the option to go and live in the wild, but instead decide to stay in the city and work until they die.

        Livestock animals on well managed free range farms have it better than humans in many ways. They are provided with food and shelter, don’t have to work, and have a quick and painless death rather then the typical brutish one they’d have in the wild.

        If I had to choose between being a wild animal, or an animal in an ethical farm, I’d pick the farm, much the same as I pick to live in society as a human.

        The question to which I’m not sure, is whether I’d choose to be a work human, or a livestock human? What if rather than have to work your entire life 40 hours per week until you retire old and decrepit, you could live in luxury and leisure until they turn you into Soylent Green at 60? Maybe it’s not such a bad deal!

  57. thesilv3r says:

    Disclaimer: I have only read the first third of the post itself, I spent most of my time playing with the model (being an Excel nerd) to understand the proposition.

    The model itself seems (in my mind) to have a critical error. It asks for your assessment of “what is the moral worth of a unit of suffering/non-suffering”, but does not seem to use these fields in the frontsheet. E.g. the formula for Human-life-year equivalents for cows is “=’Animal Lives’!F136*’Animal Consciousness’!L12” but it should be “=’Animal Lives’!F136*’Animal Consciousness’!L12*’Animal Consciousness’!L24” to factor in how much I care about a cow suffering, and probably should change even further to factor in a unit of non-suffering.

    Once I corrected for that my $ equivalent costs were much more realistic. Not sure if this was corrected for in tables used in the article and the public version of the spreadsheet is the wrong version, or if I am misinterpreting how it should be used.

    • thesilv3r says:

      Edit: never mind, finished tracing everything through and the model makes sense. Disregard my comment. Something still seems iffy with my valuations but for the time being I’ll assume that’s a problem with my priors while I continue to analyse.

      • thesilv3r says:

        Yep, definitely a problem with my inputs. When I more accurately input my moral worth of a unit of suffering. I reframed in my head to be “if I had to break a person’s leg, what level of empathetic negativity would I feel vs if I had to break a cow’s leg?” or “how many cows legs would I break before I broke another human’s leg?” (assuming I have never met neither the cow nor the person before). Once calculated that way then I got much more accurate results.

        It was mostly a problem with anchoring on the defaults of 50%, 80%, etc.

        • Froolow says:

          (I’m the vegetarian collaborator)

          Just to let you know there is unequivocally a mistake I couldn’t fix, to do with the way discounting is applied through the model – I would normally use a Markov chain (or VBA with the same effect) to implement discounting, but since I wanted all sensitivity analysis to be carried out with formulas so nobody was worried about me executing arbitrary code on their computer I had to use a bit of a kludgy approximation.

          Therefore, increasingly extreme values in any parameter that discounts will gradually start producing some strange results, but I think it should be reasonably robust over plausible entries. If you (or any other Excel nerd) knows a better way to implement that outcome I’d be really grateful if you could share it, as I wasted *hours* trying to get it to behave!

  58. hnau says:

    For reasons specific to my own point of view I wasn’t happy with this collaboration’s approach, but in most other ways it was very impressive. Things I especially liked:
    – Willingness to do Fermi estimates of weight / impact, making clear which considerations dominate the decision
    – Documentation of model and analysis of how different weights / assumptions affect the overall calculus
    – Serious effort to account for substitution effects
    – Neat approach to analyzing utility cost of impacts on the environment– I’d never seen that before
    – Engaging writing / presentation style

    It took me a while to realize that the analysis takes a rather specific set of meta-ethical assumptions for granted: “positive” utilitarianism, consciousness / preference for life as a basis for utility, and QALYs as a utility measure. I assume this is a common (though not settled?) framework among rationalists, and I support the approach of working the problem out in a consistent way in a single ethical system. But it would have been helpful to me if the write-up had explained those assumptions up front.

    (Sidenote: I don’t accept those assumptions. The most legible reason I can give is that they seem extremely non-robust. New information, reframing, analysis of second- and third-order effects, etc. can all shift the outcome dramatically between an imperative in one direction and an imperative in the other, with little room for a more cautious conclusion. For example: while reading this I wondered what the authors would think of hunting e.g. deer for meat. My guess is that they would still think it wrong, but for exactly the inverse of the reason that they oppose factory farming: namely that the deer would have experienced substantial positive utility that the hunter destroys by killing it, which outweighs the utility value of the venison to the hunter. Or do population dynamics mostly cancel out this effect given responsible hunting practices? I have no idea how to tell, and that’s kind of my point.)

    Other than that, the only thing about the write-up that disappointed me was that it traded heavily on disgust reactions (graphic descriptions of living conditions) to make its points re: animal welfare. That seemed odd given how hard it was trying to be rational / utilitarian in other respects. It made me wonder if the strongly vegetarian collaborator was the one driving the framing and presentation.

    On the other hand, despite my unwillingness to accept the collaboration’s framework it was informative and in some ways convincing. It increases my inclination to substitute from chicken to beef/pork, from bird/mammal meat to fish meat, and from all meat to vegetarian proteins on the margin.

    I rate this collaboration as an 8/10 (no particular scale or judgment implied, this is just for my own reference). As always, many thanks to the authors for putting in the work to create this.

    • acymetric says:

      Agreed. I think this is the best collaboration so far, but neither position is especially close to my own so I didn’t feel I actually took a whole lot from it personally.

      Also agree with what has been mentioned elsewhere, that none of the collaborations so far have felt particularly adversarial (it seems like everyone started off not that far off from each other).

  59. Jakub Łopuszański says:

    Given that this is a collaboration I don’t understand why many of sentences are written in first person.

    • hnau says:

      Probably hold-overs from an earlier draft written by a single collaborator. They caught most of them but there are still a few “I”‘s and “my”‘s left.

    • Froolow says:

      (I’m the vegetarian collaborator)

      We originally wrote the collaboration entirely in the first person, to make it read more like a ‘typical’ SSC article and to make it clear that every sentence was agreed by both parties such that we could speak with one authorial voice. But everyone we sent our final draft to absolutely hated that, so we did a last-minute find and replace to switch to the third person and must have missed some!

  60. HaraldN says:

    Very good post. Started feeling quite bad about my meat eating in the middle, but after some googling it turns out Sweden has very strong laws in this area, and arguably (I’d like to do more research and become more confident) winds up firmly on the side of “all food animals in Sweden prefer being farmed to nonexistence”.

    So at least I’m only hurting myself and the environment.

  61. omegaile says:

    This is an actual wooly cow.

  62. Jack V says:

    It felt like addressing circumcision in terms of “does it reduce or increase harm” is sensible, since I think that’s the bit that’s most uncertain, and if people have varying value systems about whether that’s the only thing to care about, or whether they care about bodily autonomy additionally, their eventual conclusions will still be clear, and can’t be helped by the essay.

    But for animal farming, it feels like in most scenarios you’re going to have a lot of animals in a “not great life but not literally the worst possible” range, with possible exceptions at both ends.

    And that’s directly a “what value system do you have” question: do you think there’s a cross over point of lives with a “zero” utility, and some are better and some are worse? do you think positive and negative utility are comparable and summable? do you think you add up the utility for all creatures? or average it? or weight it by sapience?

    As far as I can tell, there’s no consensus on any of those questions, so I don’t know how the animal farming question can ever come to a conclusion.

  63. Dean says:

    “the realistic alternative to meat eating is not a better life but for those animals to not exist in the first place.”

    This seems like a weird counterfactual assumption… If everyone went vegan tomorrow, a lot of land currently used for grazing or growing grain for animal feed would suddenly be up for grabs. It seems likely that a lot of it would return to wilderness, so there would be wild animals living on it. It seems the more relevant comparison is whether the lives of wild animals is better or worse than that of industrially farmed animals.

  64. Aapje says:

    Some comments:

    – The collaboration seemed very US-centric (ag-gag laws; not being aware that sheep typically roam outside in the EU, so it’s not just cows that do so; no mention of horse meat, etc).

    – There was pretty much only attention for the negative aspects of animal (factory) farming, not positive aspects. This is like arguing that humans are much worse off than hunter-gatherers by only looking at the aspects where we do worse (we don’t get enough exercise anymore), while completely ignoring the many ways in which we do better (central heating!). Cows going outside is presented as giving a high quality of life, but there was a study where milk cows could go outside as they wished (from a fairly animal friendly barn), where they preferred to stay indoors 46% of the time. So unless the cows greatly favor diversity, which may definitely be true, the quality of life between a nice barn and the pasture may be small.

    – A lot of statistics or their framing feel dodgy/debatable. For example, 3.3% of chickens dying during 6/7 weeks is presented as being very high. However, human child mortality is higher than this in many 2nd and 3rd world nations, including India (second most populous country). So should we favor a ban on procreation in India before we want to stop new chickens from being bred? What is even a reasonably mortality rate for chickens? How does chicken mortality compare to human mortality, when they don’t have the same natural life expectancy, we don’t have similar levels of medicine for animals, etc. Perhaps 3.3% is extremely reasonable??

    – Too many very unclear and fairly manipulative words like ‘common.’ Debeaking is called common in broiler chickens, but Wikipedia says that debeaking is not routinely done for commercial broiler chickens. For turkeys, it seems much more common in the US than elsewhere in the West and it was developed in the US, both of which suggest that it may be more common in the US. In some countries, debeaking is illegal. So what does ‘common’ mean? I still have no clue what the overall percentage of debeaking is, nor the percentage of the chickens sold in the US, nor sold in the EU, nor sold elsewhere.

    – I found the (attempts at) humor to be distracting and to create a sense of unseriousness.

    • Elena Yudovina says:

      Isn’t India’s child mortality rate 3.3% per (first) year, rather than per 6-7 weeks? Scaling up the 3.3% per seven weeks gets you 22% mortality for the first year of life. Casual Google searching for infant mortality in hunter-gatherer societies or medieval Europe suggests that you see this ballpark of mortalities there, so maybe switch your example to that?

      • Aapje says:

        It is the rate for children under 5 actually.

        However, chickens mature faster than humans (especially breeds that get used for meat), so I don’t think it makes sense to treat the time the same.

        Ultimately, chicken are not humans, so how can we compare these things in general? Quahog clams can live to be 500 years old, does that mean that humans are suffering from something that severely depresses our lifespan?

  65. Nate the Albatross says:

    I have a question: they indicate chickens raised in the EU are raised in these two ways, but in France there are two types. One looks like a US chicken. The other has much longer legs, much more dark meat and is labeled as “farm raised/cage free.”

    I’m not sure these chickens fall into the shed or broiler category. Does anyone have relevant information?

  66. ACC2019 Omnivore says:

    @Said Achmiz, @Harmless Frog, others who feel the pro meat eating side was underrepresented:

    1) I’m allowed to change my mind, and my vegetarian collaborator did change his mind on key issues. Our final report is substantially different from what either of us would have written independently.
    At the start of the collaboration, my affirmative case was that
    A) meat eating was crucial for development of a larger brain and is nutritionally much more bioavailable than other food sources
    B) ‘all animals are conscious and capable of suffering’ is naive and ignores the neurological/behavioral evidence for how wrong our intuitions are about the nature of consciousness
    C) the environmental/nutritional science in favor of vegetarian diets is low quality and selectively presented
    D) typical vegetarian intuitions about animal welfare don’t take into account the habituation literature which I find highly persuasive, and that if animals prefer existence and the counterfactual to a farming method is non-existence, then it is not a net harm.
    A: I determined fairly quickly that the evolutionary case for meat is primarily about caloric density and consumption speed of raw food, so I cut this section. I made my case that animal protein is important for nutrition in infants and those unable to afford a diverse vegetarian diet. If we hadn’t focused on meat eating in the west, this section would have been much longer. However, I was persuaded that in developed countries nutritional deficiencies are not a major concern.
    B: I am happy with how the section on consciousness turned out, and am glad that some found it useful. I personally updated my estimate of the likelihood of chicken consciousness significantly upwards, from about 40% because of recent discoveries of similarities between the mammalian cerebral cortex and the corresponding dorsal ventricular ridge in birds, and after reading the extensive behavioral literature on their intelligence. Conversely, my collaborator didn’t provide me with evidence that fish are conscious or capable of suffering that I found persuasive or wasn’t directly addressed by my literature.
    C: Sometimes, one has to come to a conclusion on complicated topics in the absence of definitive randomized controlled studies. I spent a lot of time going through various environmental/health papers, and I continue to have low confidence in our conclusions because we don’t know why vegetarian diets are better or what the cost of environmental cleanup will be under future technologies that may or may not be developed, but I found the totality of the evidence persuasive and quite far from my initial priors.
    D: I believe that the habituation and happiness literature is strong justification for a pronatalist position, and that remains our strongest disagreement. For example, my collaborator continues to think factory farmed pig lives are significantly worse than non-existence, I think they are slightly better, and in the model we agreed to treat them as having positive welfare impact – that was another significant concession on his part. I convinced him that factory farmed cow lives are worth living, and he convinced me that factory farmed chicken lives are definitely not. Because of how the math works out, chickens ended up dominating the analysis under our framework.

    2) I recognize that you have different moral frameworks than we agreed to, or that you would have made different arguments in favor of meat eating. This is a serious critique that I accept. My only response is that I am not the right person to represent your position, and that we did our best to provide a flexible model that could accommodate different moral frameworks or factual conclusions. We tackled a broad topic, and stuck to questions of fact that could in principle be resolved.

  67. ACC2019 Omnivore says:

    @Said Achmiz, @Harmless Frog, others who feel the pro meat eating side was underrepresented:

    1) I’m allowed to change my mind, and my vegetarian collaborator did change his mind on key issues. Our final report is substantially different from what either of us would have written independently.
    At the start of the collaboration, my affirmative case was that
    A) meat eating was crucial for development of a larger brain and is nutritionally much more bioavailable than other food sources
    B) ‘all animals are conscious and capable of suffering’ is naive and ignores the neurological/behavioral evidence for how wrong our intuitions are about the nature of consciousness
    C) the environmental/nutritional science in favor of vegetarian diets is low quality and for political reasons selectively presented
    D) vegetarian intuitions about animal welfare are wrong because they doesn’t consider the habituation literature which I find highly persuasive, and that if animals prefer existence and the counterfactual to a farming method is non-existence, then it is justified.
    A: I determined fairly quickly that the evolutionary case for meat is primarily about caloric density and consumption speed of raw food, so I cut this section. I made my case that animal protein is important for nutrition in infants and those unable to afford a diverse vegetarian diet. If we hadn’t focused on meat eating in the west, this section would have been much longer. However, I was persuaded that in developed countries nutritional defficiencies are not a major concern.
    B: I am happy with how the section on consciousness turned out, and am glad that some found it useful. I personally updated my estimate of the likelihood of chicken consciousness significantly upwards, from about 40% because of recent discoveries of similarities between the mammalian cerebral cortex and the corresponding dorsal ventricular ridge in birds, and after reading the extensive behavioral literature on their intelligence. Conversely, my collaborator didn’t provide me with evidence that fish are conscious or capable of suffering that I found persuasive or wasn’t directly addressed by my literature.
    C: Sometimes, one has to come to a conclusion on complicated topics in the absense of definitive randomized controlled studies. I spent a lot of time going through various environmental/health papers, and I continue to have low confidence in our conclusions because we don’t know why vegetarian diets are better or what the cost of environmental cleanup will be under future technologies that may or may not be developed, but I found the totality of the evidence persuasive and quite far from my initial priors.
    D: I believe that the habituation and happiness literature is strong justification for a pronatalist position, and that remains our strongest disagreement. For example, my collaborator continues to think factory farmed pig lives are significantly worse than non-existence, I think they are slightly better, and in the model we agreed to treat them as having positive welfare impact – that was another significant concession on his part. I convinced him that factory farmed cow lives are worth living, and he convinced me that factory farmed chicken lives are definitely not. Because of how the math works out, chickens ended up dominating the analysis under our framework.

    2) I recognize that you have different moral frameworks than we agreed to, or that you would have made different arguments in favor of meat eating. This is a serious critique that I accept. My only response is that I am not the right person to represent your position, and that we did our best to provide a flexible model that could accommodate different moral frameworks. We tackled a broad topic, and stuck to questions of fact that could in principle be resolved.

    • Froolow says:

      (I’m the vegetarian collaborator)

      For those who believe the collaboration is one-sided, I think you are probably working with a mental model where a middle-of-the-road vegetarian and a middle-of-the-road omnivore had a discussion and ended up agreeing the vegetarian was completely right. In fact, my prior belief could probably best be described as ‘raving’, so the fact the collaboration ends up concluding a sort of middle-of-the-road vegetarian position means neither side got to put their views across unopposed, and probably I moved further from my prior beliefs than my partner did.

      Scott asked us to write these entries in a more collaborative style than last year, so perhaps the ‘behind the scenes’ disagreement isn’t obvious, however from my point of view the meat eater ‘won’ the debate, if winning is judged as moving less far from your starting point / conceding fewer points.

      ACC2019 Omnivore is too modest to say so, but he substantially altered my opinion on a lot of topics. A partial list of some absolutely fundamental points I changed my mind on:

      1) Fish probably aren’t conscious (or at least, we currently lack the evidence to prove they are)

      2) It is likely better to eat cows than chickens (which I didn’t believe when the question did the rounds on SSC a year or so ago, but now do believe thanks to very strong and specific arguments from my ACC partner)

      3) There is no strong signal that vegetarianism is cheaper than meat eating (especially if you need to consume more to compensate yourself for lost enjoyment)

      4) There are some at least reasonably plausible circumstances under which the harms of eating meat result in trading off your own for social utility, rather than always being net beneficial for both yourself and society

      5) Animal intelligence is interesting but is not the same thing – or even necessarily a proxy for – animal consciousness, which is the morally relevant issue

      • Said Achmiz says:

        For those who believe the collaboration is one-sided, I think you are probably working with a mental model where a middle-of-the-road vegetarian and a middle-of-the-road omnivore had a discussion and ended up agreeing the vegetarian was completely right. In fact, my prior belief could probably best be described as ‘raving’

        No, your supposition is entirely wrong. As I say in my first comment:

        The impression I got was that one person was an avowed vegetarian and the other person was just some vaguely-“rationalist” sort who ate meat and thought that eating meat was okay but didn’t really have any strong opinions on the matter.

        Your comment here, and that of your collaborator, entirely confirms that impression, and thereby also confirms my judgment of one-sidedness.

        Finally:

        the fact the collaboration ends up concluding a sort of middle-of-the-road vegetarian position means neither side got to put their views across unopposed, and probably I moved further from my prior beliefs than my partner did

        You must realize, of course, that this is a classic and well-worn tactic: begin with a position that is far more extreme than the one you’re aiming to have the collective (community, society, etc.) end up with; allow your opponents to force a compromise; end up with a position that’s still well on your side of the “center”. Note that I am not accusing you of any conscious (or even, really, unconscious) deviousness; I am only pointing out that your posited conclusion (in the quoted bit) is entirely unwarranted.

      • Said Achmiz says:

        Here is an interesting side note, which further illustrates the degree to which this AC has managed to miss a broad swath of views in the space of possible views on this subject. You say:

        A partial list of some absolutely fundamental points I changed my mind on:

        2) It is likely better to eat cows than chickens (which I didn’t believe when the question did the rounds on SSC a year or so ago, but now do believe thanks to very strong and specific arguments from my ACC partner)

        Now, for myself, I do find certain arguments in favor of reducing certain sorts of meat consumption to be compelling. (These are all based on impact on humans, of course.) I have, in fact, mostly eliminated my consumption of certain meat—namely, pork (I don’t eat very much beef either, but this is due more to personal preference than any sort of conviction).

        And one thing that is clear to me is that chicken is the very last sort of meat that it makes any sense to reduce consumption of, and the most efficient sort of meat, in the sense of producing the greatest benefits for humans while incurring the least negative consequences for humans. (And I noticed that you did not discuss the relevance of income level or cultural background to consumption and meat selection—two factors which I consider critical to this issue.)

        Consider, then, that people who have precisely my moral views may, indeed, be convinced to eat less meat—but they will be convinced by arguments of which you (so it seems) are not even peripherally aware!

        If I haven’t said so before, let me say this now: I appreciate the amount of work you (both of you) have put into this collaboration. Read my criticisms only as expressions of regret for what could have been, and dismay at the unfortunate consequences of so well-intentioned an effort.

    • Said Achmiz says:

      I recognize that you have different moral frameworks than we agreed to, or that you would have made different arguments in favor of meat eating. This is a serious critique that I accept. My only response is that I am not the right person to represent your position

      Fair enough. I find it regrettable, however, that such collaborations are undertaken in the first place. I fear that the result will be that people will read this sort of thing, and come away with the impression that the matter was seriously debated and investigated, and both sides truly represented, and “meat eating is bad on net” was concluded (when in fact, as far as I am concerned, one side did not even show up). I would have liked to see a strong, and prominently displayed, disclaimer, to the effect that you do not represent one major class of arguments against vegetarianism, and that one of the major positions in this debate has not received any representation in this AC. Otherwise, for all your quite admirable effort, I cannot help but see this entire exercise itself as net-negative.

      we did our best to provide a flexible model that could accommodate different moral frameworks

      This, however—I regret to say, and hope you will not take offense at my saying—is not a credible claim. Surely you could have done better, and easily. (Why, your very spreadsheet seems to permit a better attempt to represent my sort of view—just set some of the values, for moral worth of animals and so on, to zero! Could you not have included that position on your chart? It would have provided at least some counterpoint to the extremist “no speciesism” option!)

      • ACC2019 Omnivore says:

        No offense taken:
        Re. your particular point about having a ‘animals have 0 moral value’ sensitivity analysis, we segmented our analysis in the table right above by animal welfare/environment/health/finance. As far as I’m concerned that analysis already exists as a subset of the table. As do all permutations of ‘I only care about category X’ or ‘You made an unpersuasive case regarding Y.”

        The included ‘chickens not conscious scenario’ is actually more pro-meat than a ‘animals have 0 moral value’ scenario because all other animals were concluded to have net positive lives after habituation is taken into account, and is almost identical in value.

        I fear that the result will be that people will read this sort of thing, and come away with the impression that the matter was seriously debated and investigated, and both sides truly represented, and “meat eating is bad on net” was concluded (when in fact, as far as I am concerned, one side did not even show up).

        Can you tell me, roughly, what you think a productive public debate about whether animal experience has moral weight might look like?

        You said earlier

        How much I care about chickens (regardless of whether or not they “suffer”, regardless of whether or not it even makes any sense to speak of such animals “suffering”, etc.), is certainly a matter of my moral views, and certainly not a matter of any external “facts”.

        If I can’t use facts to persuade you (which is completely fine), then I don’t understand your objection, because it would seem to apply only to someone who is persuaded by our conclusion without reading what we’ve written. I guess I’ve only seen appeals to authority mislead on questions of fact – when it comes to moral persuasion they work approximately never.

        Finally, you say

        I would have liked to see a strong, and prominently displayed, disclaimer, to the effect that you do not represent one major class of arguments against vegetarianism, and that one of the major positions in this debate has not received any representation in this AC.

        We wrote at the top that

        The central question is whether factory farmed animal lives are worth living….and finally, we survey the more easily measurable impacts of meat eating on environment, finance, and health.

        If these are areas that a particular reader finds irrelevant to the question, I’m not sure how they would end up being misled by our conclusion.

        • Said Achmiz says:

          Addressing only this part for the moment:

          We wrote at the top that

          The central question is whether factory farmed animal lives are worth living….and finally, we survey the more easily measurable impacts of meat eating on environment, finance, and health.

          This part of your write-up reads as saying that “whether factory farmed animal lives are worth living” is the central question of the whole issue at hand—and not that that’s the central question of your investigation (and that there may be other issues, perhaps even more central to the whole question, which you chose not to address). Was the latter the intended interpretation? If so, this was not at all evident. If, on the other hand, the intended interpretation was the former, then my criticism stands.

          • ACC2019 Omnivore says:

            Ah, yes. I now understand your objection. Thank you for clarifying. No, we were saying that the experience of animals is the most relevant part of the debate, providing a roadmap for our discussion. I agree that we don’t make the claim ‘we did not consider X,’ such that if someone did think any particular X was relevant but we never addressed it, they could, having not read the essay, get the false impression that X is considered and judged to be less important. Since your position is there is such an X, then we are now on the same page.

        • Said Achmiz says:

          Re. your particular point about having a ‘animals have 0 moral value’ sensitivity analysis, we segmented our analysis in the table right above by animal welfare/environment/health/finance. As far as I’m concerned that analysis already exists as a subset of the table. As do all permutations of ‘I only care about category X’ or ‘You made an unpersuasive case regarding Y.”

          In a sense, I suppose that’s so. It’s somewhat hidden, and not discussed, and absent from the subsequent chart. (By the way, speaking of that table you refer to: you calculate that, taking into account impact on animals, the total cost of eating meat is the equivalent of less than $10k over a lifetime… and you summarize this as “Overall, the impact of eating meat like a typical person is likely to be substantially negative.”! What the heck?!) To be frank, “someone could carefully read our data and arguments, and extract from them, after consideration, some conclusion about some view” is not the same thing as “we, ourselves, bring up and analyze that view”. What you describe as a “sensitivity analysis”, I see as something that makes the majority of your write-up simply irrelevant! This isn’t some small quibble about data presentation…

          Can you tell me, roughly, what you think a productive public debate about whether animal experience has moral weight might look like?

          Why should this be what’s debated? No, the productive public debate would be: given that many people don’t consider animals to have moral weight, what should be their views about eating meat? What is the best way for such people, and avowed vegetarians, to cooperate—are there any aligned goals? (This might include a much more serious discussion of environmental impacts, and of cultural and income-based considerations, etc.) In what way can the two groups’ interests be harmonized? (This might include a discussion of hunting, non-industrial-scale farming, lab-grown meat, etc.) And so on.

          If I can’t use facts to persuade you (which is completely fine), then I don’t understand your objection, because it would seem to apply only to someone who is persuaded by our conclusion without reading what we’ve written. I guess I’ve only seen appeals to authority mislead on questions of fact – when it comes to moral persuasion they work approximately never.

          I confess that I don’t follow what you’re saying here. Assuming that the previous paragraph doesn’t obviate your point in this one, I would ask that you try to rephrase or explain what you mean here.

          • ACC2019 Omnivore says:

            you calculate that, taking into account impact on animals, the total cost of eating meat is the equivalent of less than $10k over a lifetime… and you summarize this as “Overall, the impact of eating meat like a typical person is likely to be substantially negative.”! What the heck?!

            These are annual numbers, as stated several times in the same section, table, and chart.

          • ACC2019 Omnivore says:

            The productive public debate would be: given that many people don’t consider animals to have moral weight, what should be their views about eating meat? What is the best way for such people, and avowed vegetarians, to cooperate—are there any aligned goals? (This might include a much more serious discussion of environmental impacts, and of cultural and income-based considerations, etc.) In what way can the two groups’ interests be harmonized? (This might include a discussion of hunting, non-industrial-scale farming, lab-grown meat, etc.) And so on.

            Ah, yes. Thank you very much, that was quite helpful. I’m sorry, that’s simply a different question than we were interested in collaborating on. Most people think animals have moral value, based on surveys covered on this blog, and so did both collaborators, albeit with large differences in magnitude.

          • Said Achmiz says:

            These are annual numbers, as stated several times in the same section, table, and chart.

            … huh? It says “Total Life Years”, and… oh, I see, you don’t mean ‘total’ like, total, you mean “total annually”. Confusing! But, fair enough, objection withdrawn.

          • Said Achmiz says:

            Most people think animals have moral value, based on surveys covered on this blog

            You mean “most people who read this blog”, surely? Or do you say that said surveys are representative of the general population? Or is it simply that your analysis and conclusions are not meant to apply outside of SSC’s readership?

            EDIT: Or perhaps I’m simply confused about what surveys you refer to. Do you have any links handy?

          • ACC2019 Omnivore says:

            In section 1.5 we linked to a couple surveys that have been referenced on this blog, neither of which were representative of general population.

            If you’re looking for something more representative, I’d point you to the first random telephone survey I found, in which a majority of people claim to agree with positions that wouldn’t make much sense if they placed 0 moral value on animal suffering (agreeing with various views such as “government should actively promote animal welfare,” “I would vote for animal welfare laws,” “animal well-being is more important than low food prices,” etc.)
            http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.567.683&rep=rep1&type=pdf

            Our analysis and conclusions are meant to apply to anyone who cares about any of the topics we covered in the collaboration: money, animal welfare, their own health, or the environment.

          • Said Achmiz says:

            the first random telephone survey I found, in which a majority of people claim to agree with positions that wouldn’t make much sense if they placed 0 moral value on animal suffering (agreeing with various views such as “government should actively promote animal welfare,” “I would vote for animal welfare laws,” “animal well-being is more important than low food prices,” etc.)

            … it should be quite obvious why this sort of thing is not even remotely a reliable indicator of people’s actual moral views.

            But, in any case, you’ve certainly answered my questions, and I can ask nothing more for now. I’m content to let this thread come to a close.

  68. xagaros says:

    Overall, this makes me more likely to become vegan. My criticism: I understand that we need a proxy for “proximity to threshold of life worth living”, but suicide cannot be a good one. As they note, slaves in antebellum America committed suicide *less than* whites or freed slaves. (I’ve previously run across this puzzling statistic in my own reading.) This is actually an error of sign, not just magnitude, and unless you think slaves had lives more worth living than free whites – which by the understandable disclaimer in the text where they cite this, the collaborators do not – then we must conclude that suicide is not a useful indicator. And arguing that it works across species is more than “stretching it”. In a grim thought experiment, we can imagine a hellish facility where we torture people more and more, but always give them the option of suicide – most of us would agree you’d see a correlation with torture severity (and therefore, worthiness of living.) Animal suicide is almost unheard of and I doubt we’d see such a correlation regardless of the torture severity, and many people would argue that caged chickens are living in exactly such a facility. As to what to replace this proxy measure with, I’m afraid I don’t have an alternate suggestion.

    Also, I could’ve sworn I saw a link once on Marginal Revolution that wild animals suffer less than those in captivity, but when I looked for it I could only find a link to this Rolf Degen retweet saying the opposite. https://twitter.com/DegenRolf/status/841320785423847424

    • dionisos says:

      Thanks.
      And I think we should also take into account we evolved through natural selection to try to stay alive, not to take rational decision about live worth living.
      It say something that we still kill ourselves.

  69. Rafal Smigrodzki says:

    For an avowed speciesist like me, being forced into vegetarianism for ideological reasons would be extremely distressing. I eat animals in part because I value my and other peoples’ lives and experiences much more, perhaps even incommensurately more, than the lives and experiences of commonly-eaten animals.

    To tell me I may not eat animals would be to equate the intrinsic (not instrumental) value of human life with the value of a gaggle of chickens, or a cow.

    Ideological vegetarianism is thus an evil, anti-human doctrine. Health-conscious vegetarianism is simply incorrect. Cost-based vegetarianism is something to be endured while getting rich and not an independent argument in favor of vegetarianism.

    I will order a tomahawk rib-eye today at my favorite steak-house, 42 oz., medium rare, with blue cheese crust. A man and his woman have to eat.

    • MichaelStJules says:

      To tell me I may not eat animals would be to equate the intrinsic (not instrumental) value of human life with the value of a gaggle of chickens, or a cow.

      You don’t have to equate their value to think that your trivial interests in eating them should not outweigh some of their most fundamental interests, e.g. in not suffering or in life itself. If you found it entertaining to burn dogs alive and no other harms would result from it, do you think that it would be ethical to do? How about infants or early toddlers instead of dogs?

      Some further reading, in case you’re interested:
      https://ea-foundation.org/blog/the-strongest-argument-for-veganism/
      https://www.animal-ethics.org/ethics-animals-section/

      • Said Achmiz says:

        Your first link is an interesting one. If that really is the strongest argument for veganism (as determined by the Effective Altruism Foundation, no less!), then I cannot but conclude that veganism has no strong arguments whatsoever for it. After all, not only is the given argument not at all convincing, it’s not even valid—it’s got easily noticeable errors of reasoning in it! And this is the best that the vegan / animal-rights folks of the EA movement can come up with? I am more convinced than ever that there’s not even a token moral case for going vegan…

        • MichaelStJules says:

          What errors are you referring to?

          Whether or not it’s the strongest argument for veganism is subjective, but I do think it’s fairly strong. A lot is hidden in the word “unnecessary”/”unnecessarily”.

          • Said Achmiz says:

            What errors are you referring to?

            The linked page says:

            Animals shouldn’t be made to suffer unnecessarily, and harming them for no good reason, i.e. without necessity, is what defines “cruelty to animals” and what civilised legislations forbid.

            Yet that “i.e.” is invalid; “for no good reason” and “without necessity” are not synonyms (though, as you say, a lot hides in this word “necessity” in its various forms). Do I we shouldn’t harm animals for no good reason? Indeed, I do; it would be cruel to inflict harm on an animal without having a good reason to do so. The trouble is that, quite simply, “I want to eat it” is a good reason. (Does it constitute “necessity”? Well, it depends on how you define “necessity”; the usual usage of the word wouldn’t include “I want to eat it”. But in that case, I don’t agree that animals shouldn’t be harmed without necessity!)

            So, you see, the logic does not work. The argument simply assumes either that “for no good reason” and “without necessity” are synonyms, which of course they are not; and equivocates on the meaning of “necessity”. Both of these are errors. (They might even be semi-deliberate sleights of hand; but I hesitate to impute so devious a mindset to the page’s author(s), and in any case blindness to argument flaws due to fanaticism is a more probable explanation than conscious deception.)

          • NoRandomWalk says:

            @Said Achmiz,

            I’d like to understand what your moral value/system is, and how common it is from your perspective.

            If, for whatever reason, your paying some amount of money would result in an animal not having their leg broken just before dying of natural causes anyway, up to what cost would you be willing to pay so that this suffering does not occur for the cases of: another human, chimp, cow, chicken? Also, what do you think the typical person would answer, for the same four values?

          • Said Achmiz says:

            @NoRandomWalk:

            If, for whatever reason, your paying some amount of money would result in an animal not having their leg broken just before dying of natural causes anyway, up to what cost would you be willing to pay so that this suffering does not occur for the cases of: another human, chimp, cow, chicken?

            Another human: depends who it is. Someone close to me? Obviously, quite a bit. Otherwise, depends on various factors.

            EDIT: You say “If, for whatever reason, …”, but that is clearly an absurd hedge—the reason matters greatly. The specifics of the situation would affect my actions (in the “human who is not someone I personally care about” scenario, mostly).

            Chimp, cow, chicken: nothing.

            Also, what do you think the typical person would answer, for the same four values?

            It depends on who “the typical person” is, I suppose, but for many such classes of “typical” people, I’m sure they would say that they would pay some non-trivial amounts of money for these things.

            Whether they actually would pay this money, were your hypothetical situation to occur, is another matter—and here I think the answer is “most people wouldn’t, in the median, pay anything for such a thing”. (Of course various factors would affect this, such as whether their action were secret or public, whether it was the first such situation or the 10,000th, and many other circumstantial things.)

          • NoRandomWalk says:

            EDIT: You say “If, for whatever reason, …”, but that is clearly an absurd hedge—the reason matters greatly.

            I agree it matters greatly. That was the point of the hedge. Without it your answer would tell me a lot less about you.

            I disagree with you about the median person. I think the majority of people would, if the offer was made in secret for the first time, pay a non trivial sum (say, $10+, 30 min of their time, etc) for any of a chimp/cow/pig to not be slowly tortured to death.

          • thisheavenlyconjugation says:

            I interpreted “no good reason” to mean “because it makes me a little bit happier” — cruelty to animal laws forbid torturing animals even if you have the reason of your own enjoyment. In other words, the argument goes “harming animals for your own pleasure from seeing them suffer is wrong (and not just for virtue-ethicsish reasons about encouraging bad habits in humans)”; “in general, getting to eat meat isn’t a significantly more compelling reason than getting to see suffering”; “therefore eating meat is wrong”. I agree that this is not actually what’s written though; you would only think to substitute this argument if it was already on your mind.

          • Aapje says:

            I interpreted “no good reason” to mean “because it makes me a little bit happier”

            This is a thoroughly anti-human position, though.

            The vast majority of what humans do is for “no good reason:”
            – We eat food more tasty than what is necessary for survival
            – We have homes bigger than we need to survive
            – We heat our homes more than we need to survive
            – We have partners and children, than we don’t need to survive
            etc, etc

            If you treat pleasure as a bad reason to do things, then it seems to me that you have to consider most of what humans do to be evil.

          • HarmlessFrog says:

            What errors are you referring to?

            This is false:

            (3) The consumption of animal products is unnecessary.

            Humans require at least some portion of their food to be in the form of animal foods, otherwise they aren’t getting several key nutrients or getting them in substandard amounts – complete protein, fat soluble vitamins and essential fatty acids, especially – and this point applies for children ten times as much as for adults. Artificial supplementation as good as the real thing doesn’t exist yet, so consumption of animal foods is quite necessary, unless one thinks that widespread nutrient deficiencies, stunting, infertility and supplement toxicity are an appropriate price to pay for never hurting an animal with consumptive intent. (Nevermind that animals would still, necessarily, suffer, as all living beings suffer.) And that’s not even getting into whether we possibly could feed our population if we remove all the planting-unsuitable lands we use for livestock from agricultural use.

  70. Robert H says:

    The “lives worth living” argument also implies that it’s morally right to raise humans for their organs like in the movie “The Island” or “Never let me go”. In these fictional cases the humans wouldn’t exist if not the market for their organs and they live good lives until their (mostly painless) death. Theoretically the can have wonderful lives and not even know that they are being breed for organs. But this regardless doesn’t seem morally right to me. Has anyone dug deeper in to this counter-argument?

  71. MichaelStJules says:

    I think it’s very likely that fish and crustaceans feel pain, and while I haven’t had the chance to scrutinize all of the arguments and evidence, several other invertebrate species seem likely to feel pain, too. See the following paper:

    Sneddon, L. U., Elwood, R. W., Adamo, S. A., & Leach, M. C. (2014). Defining and assessing animal pain. Animal behaviour, 97, 201-212.
    https://www.researchgate.net/publication/266794956_Defining_and_assessing_animal_pain

    And the more recent and comprehensive (but not formally peer-reviewed) report by Rethink Priorities on invertebrate sentience:
    https://www.rethinkpriorities.org/invertebrate-sentience-table

    My impression is that behavioural and cognitive indicators (e.g. selective attention, motivational tradeoffs, depressive-like behaviour, play-like behaviours, other emotions and social behaviours) are given too little consideration and weight compared to neural ones.

    Some popular coverage of some research for fish:
    https://www.nytimes.com/2017/10/16/science/depressed-fish.html
    https://www.independent.co.uk/news/science/fish-sentient-animals-friends-positive-emotions-study-study-source-ethics-eating-pescaterians-vegans-a7660756.html

    • MichaelStJules says:

      To add to this, if an individual displayed all of the behaviours of a fully conscious typical adult human, but their brain was very different from ours and poorly understood, I would still assume they’re conscious. I’d expect that they have some kind of neural architectures doing similar things to our own brains, but perhaps very differently organized, and even if we haven’t identified them precisely.

      I think relying too much on comparisons between very different brains will mislead us. Brain regions with similar developmental and evolutionary origins need not have the same functions. If an individual has a given capacity that we think is indicative of consciousness, how much does the underlying mechanism actually matter?

      Also, see responses to the cited article about fish:
      https://animalstudiesrepository.org/animsent/vol1/iss3/

      • sty_silver says:

        In general, if a brain is similar to a human brain then that’s strong evidence for consciousness, but if it’s not that certainly doesn’t prove the absence of consciousness. It’s technically evidence against consciousness in the Bayesian sense, but only because the particular argument of consciousness-by-similarity is missing. The absence of one argument for X doesn’t prove X is false. It’s completely possible that a very different brain also leads to (a very different?) consciousness.

        Not saying anyone claimed otherwise; I haven’t looked at the sources. It’s just a confusion that I suspect might come up.

  72. Spiritkas says:

    Thanks for putting this together. As an ex-vegan/ex-vegetarian I think you gave the nutritional aspect far too little weight. Partly there is a lot of new science and partly there are not many twin studies to see the health impacts of lower lifetime levels of minerals, vitamins, and essential fats. There is also the time of transition where healthy bones might form with meat eating for those who transition as teens or adults vs infants. I’d think this would be more possible to study with some populations having childhood/infant lifelong vegetarianism vs others who switched later in life.

    Those points aside, there are several key nutrients you did not consider and a keyword search of your article and the comments tells me this has not come up yet. Creatine and Choline. The body is able to create a minimum amount of creatine, but the effects of a dietary deficiency are unclear. Certainly it affects muscle performance and any vegan athlete will be supplementing with it. But what about a person leading a regular lifestyle with only moderate to, let’s be honest, negligible exercise? Just because we don’t currently have a super clear cut answer to a question doesn’t mean it isn’t’ worth considering as a possible source of harm.

    Choline is far more clear cut and even people who eat meat are deficient in this often times. Due to the lack of organ meant in our diets. I’d recommend you look into this one quite a bit further to see the health effects on the brain and cellular integrity. Deficiency here is very common and does have real health effects, but the topic of choline has a low public profile. I’d be interested in looking at neuromuscular issues in vegans and vegetarians.

  73. Rand says:

    I was very confused about the absence of eggs in this article.

    The article says that “Broadly speaking, caged chickens have literally no human analogue in terms of how much they suffer.” But caged chickens are not considered in this article – because they aren’t eaten in the US? Only used for eggs?

    I had to reread several times before I realized the authors were laser-like focused on vegetarianism, to the exclusion of veganism. (Despite the one graph with veganism in it.) Having gone to all this work, this seems like a crying shame. This is doubly true if the actual conclusion of this article is “eat meat, don’t eat chicken and for the love of all that is holy, don’t eat eggs”? And that’s what it sounds like, especially given that caged egg chickens suffer for longer per calorie of protein (1.4x longer, according to https://reducing-suffering.org/how-much-direct-suffering-is-caused-by-various-animal-foods/.) (Though maybe habituation plays a big role here and I don’t know how much eggs vs. chicken most Americans eat.)

    Is that basically it?

    (Now I’m also curious whether the “vegetarian” author eats eggs and how they square that with their conclusions.)

  74. Robert Beckman says:

    Subtle error.

    In the testosterone link they found that both SHBG and total testosterone were slightly increased, but not free testosterone – which has more virility effects. I’m a typical case an increase (in males) of SHBG and total testosterone by the same amount would result in lower free testosterone as it’s bound up by the SHBG. The actual endocrinology is more complicated, but in short their finding means the opposite of what you implied – largely because the abstract is poorly written (either by someone who didn’t understand what it meant, or someone knowing they were talking to fellow experts and didn’t need to clarify).

  75. danielszev says:

    Information about the animal conditions. I shop at costco so that’s what I researched. Also they seem like a good typical example (citation needed).

    Chicken: (broiler) the article spent most of its time looking at cage animals which most of Canada, US is not. If this wasn’t a SSC article I would have suspected the pics were trying to get an emotional response, over conveying information. I shop at costco so that’s what i looked up. Costco uses the standards from this site https://www.costco.com/sustainability-animal-welfare.html (evidence they use the standards) https://www.chickencheck.in/chicken-guarantees/ (the standards). Poke around the site if you actually care to learn about it.

    For those not wanting to read it (don’t worry there are sfw pictures) the spaces do seem cramped, but it doesn’t look like anything that posted in this study. If you wanted actual information on how they are being kept this is where you would look.

    Pigs: They are phasing out the gestation cages and will be done with them in 2 years. (source) https://www.costco.com/sustainability-animal-welfare.html
    Castration without anesthesia looks like this https://www.youtube.com/watch?v=qDvoetuy6DI skip to 6:53. Not saying he enjoyed it but it doesn’t look hes suffering that much.

    Costco follows the “Pork Quality Assurance Plus program” (sic). This is that programs site https://lms.pork.org/ and a guide to housing with pics https://porkcdn.s3.amazonaws.com/sites/all/files/documents/2013SowHousingWebinars/1%20-%20Group%20Housing%20Systems.Choices%20and%20Designs%20-03643.pdf

    Pretty cramped, reminds me of prison, but still space to move around.

    Got bored before I did cows.

    The thing I like about this study is that it touches upon most of the relevant points. I would love to see this as a jumping off point for further fleshing out this issue.

  76. brownbat says:

    I found the “better than nonexistence” bar weird, though I get how you got there.

    I’m not automatically convinced I have a moral obligation to sustain the existence of other lines of animals as the authors do, because I worry this would lead quickly to confusing odd results. (How much effort should we devote to selectively breeding greater happiness into as many mammals as possible and actively culling the rest? Maybe deer are terrified most of their lives because of the constant threat of predation, unaided by the lack of access to veterinary care, and we should be aggressively eating as much venison as possible?)

    Also, it seems you shouldn’t be allowed to create conditions that make life not worth living and then have clean hands when calculating whether that thing should exist or not. I think a better approach would be to separate the moral questions out into discrete parts.

    First, would that animal choose the natural state, with predators and disease and lower access to food, but greater freedom of movement?

    Second, what steps ought we take to maximize valuable life year experiences on the planet? (Maybe we have some obligation to replace all farms with petting zoos filled with happy dogs and capuchins.)

    You might separate out as a third question whether or not adding the threat of predation to particularly sapient species has any moral weight. I suspect it does, but I can imagine attacks on this view (e.g., what if the animal has no idea that’s what it’s being used for, so there’s no impact on subjective experience?) I think it’s wrong to play russian roulette with a sleeping friend, even if the gun fails to fire. Maybe with predation you’re engaging in behaviors that risk creating existential dread, even if they are unlikely to do so in any particular case.

    FISH AND CHICKENS

    The biggest divergence for me was that after reading the fish study, I became very skeptical of using behavior as a measure of consciousness. After that, I found the linked chicken summary deeply unpersuasive.

    I feel like chicken capacity for cruelty might be a relevant moral consideration too. I’m generally a utilitarian, but given a choice between a world where harms are randomly distributed, and one where harms predominantly affect bastards, I strongly prefer the latter. I don’t see that as dispositive, but it might at least be a tiebreaker in hard cases.

    I would have also enjoyed a detailed discussion of cephalopod intelligence. Though they aren’t farmed, it would still be relevant to the question of whether or not we should be a source of violent deaths to produce calamari.

    So…

    I started as a meat eater who sometimes avoids eating pigs and squid (and would avoid eating monkeys or dogs if it was a common option), primarily driven by the desire to not be a predator to sapient species, since that generally feels like a dick move.

    I increased my aversion to pigs based on the probable unpleasant experience of farming, and shifted a bit towards quality of life as a primary concern rather than predation. Cows seem less intelligent than pigs and have better farmed experiences, so they still don’t make my threshold of concern, but they probably lead to worse climate and health outcomes of all the meats, so maybe I’ll avoid them except in the most optimal conditions — like the perfect wagyu ribeye or something.

    So after this, I expect to restrict more strictly to chicken and fish, and alternate more aggressively towards fish as a hedge against the possibility I’m undervaluing chicken sapience.

    I’d like to see more health research that differentiates between meat types. Smoked beef leads to significantly higher incidences of cancer, are preparations and meat type dominating the measured health effects? Also, some types of proteins are hard to get for vegetarians, maybe a hedge strategy with a piece of baked chicken or sushi once a week ends up being better for health than total conversion to veganism? I don’t know, would still like to see more research.

  77. JonahB52 says:

    While this subject was the least I’m generally interested in, I thought this piece was the most informative, most entertaining, and most compelling. I do take issue with sheep being considered woolly cows, as sheep aren’t finished in feedlots like cows are, but live their whole best lives in pastures ( http://agriculture.vic.gov.au/agriculture/livestock/sheep/specialist-forages-lamb-finishing-guidelines )