I wrote this as a quick explanation of why I value non-human animals the way I do. It’s not particularly thorough, and my explanation has some clear holes; this is just a general outline.

When we’re considering charitable interventions that help animals, it’s important to have some sense of how valuable it is to help those animals, which means we want to know how sentient they are.

How sentient an animal is–that is, how strongly it experiences pleasure and pain–almost certainly relates to how its brain works. I see four reasonably plausible ways that sentience could relate to brain size:

  1. Suffering is caused by certain fixed brain structures, and for certain types of physical pain (like what chickens experience on factory farms), humans and chickens have the same brain parts and therefore experience this pain equally.
  2. Sentience is linear with brain size.
  3. Sentience is sub-linearly related to brain size; for example, sentience may be logarithmic with brain size.
  4. Less intelligent animals are generally more sentient because they “are [not] capable of intelligently working out what is good for [them], and what damaging events [they] should avoid”, so they need a stronger pain response to compensate.

These are not the only possibilities, but I believe this covers the most plausible ones. (See this comment.)

The only one of these where humans and factory farmed animals differ substantially is if sentience is linear with brain size. Chicken brains are about 1/300 the size of human brains, so it’s plausible that chickens experience suffering 1/300 as much as humans. When talking about factory-farmed land animals I generally talk about chickens because there are more chickens than cows or pigs, but factory-farmed cows and pigs have greater total brain mass than factory-farmed chickens (by a factor of 10). If we account for cows and pigs too (and ignore other farm animals for simplicity), then we should adjust the value of one vegetarian-year downward by a factor of about 200 relative to if all animals (and humans) were equally sentient.

I believe the linear-with-brain-size case is somewhat implausible. Humans have big brains, but most of these brain structures don’t seem to have anything to do with pain or emotions. We have tons of brain space devoted to memory, recognizing faces, and spatial navigation, and none of these directly relate to sentience. I don’t want to make any strong claims about how consciousness works because we don’t really know, but I don’t believe these brain parts have much to do with sentience, so we shouldn’t say that humans are more sentient because we have them.

It’s plausible to me that pigs are more sentient than humans, but I find it unlikely that chickens are more sentient, because their brains are simply much less sophisticated. Pigs have fairly complex emotions similarly to humans. But again, I’m not particularly confident here; perhaps physical pain is really simple and doesn’t require much brain space.

We can more specifically pinpoint the expected sentience of non-human animals by assigning probabilities to different explanations of sentience and taking the weighted average. On this approach, it appears at first that chickens have a similar expected sentience to humans (assuming you don’t assign a tiny probability to possibilities 1, 3, and 4 above), but this suffers from the two envelopes problem, and I don’t know how to resolve this.


Brain size data taken from Carl Shulman’s “How is brain mass distributed among humans and the major farm animals?”, which pulled from several sources.

Numbers on animal-years saved are reverse engineered from Animal Charity Evaluators’ impact calculator.