We can wrong other sentient beings in three ways: by reducing their happiness or causing them to suffer (moving them down the happiness scale), by violating their rights, or by exploiting them, which means using them as mere means to our ends. I am not convinced that the second and third kinds of wrong apply to animals as they do to human beings. But we cause much suffering in animals–for example, through factory farming and the destruction of habitats–and we are obligated to address that. Reducing the consumption of meat is obligatory to the degree that it affects the supply of factory-farmed animals. But it is not the case that any killing or eating of animals is immoral. Those are my current views, and I will try to explain below.
I think about the relationship between happiness and rights in the following way. You should not cause me to suffer or reduce my happiness. But instantly killing me would not harm me in that way. I’ve had a happy life, and you would be freezing my current happiness score at its high net level. I would then suffer no more. Yet obviously you would have violated my right to life, which must be different from my interest in being happy. Why do I have a right to life? Mainly because I have plans that make sense of my actions. By suddenly killing me, you ruin my plans and make many of my past actions pointless. You also harm other people and violate their rights by removing me from their lives (or so I hope they would feel).
Now, if our beloved dog suddenly and painlessly died, his long-term plans would not be frustrated, and his recent actions would not be rendered meaningless. He has plans, such as stealing the treats out of the closet and snuggling with his human companions; but these plans are short-lived. His past treats and snuggling sessions would still represent successes even if his life suddenly ended. We would be sad, and you would violate our rights if you took him away. But I am not convinced his rights would be affected.
Likewise, our dog would be very sad to lose me and my family; suddenly killing us would cause him harm. But if this happened while he was with his dog-sitter, whom he loves, he would not be sad. The ties among animals, although profound, only matter morally insofar as they cause happiness or suffering. In contrast, human relationships give our actions purpose, and thus wrecking other people’s relationships can violate their rights even if they aren’t unhappy about it.
As for exploitation, this also violates other people’s rights because it frustrates their plans or substitutes our plans for theirs–even if it causes them no unhappiness or indeed makes them happier. I am not convinced that this concern applies to dogs and other mammals. Whether our dog is happy is the issue, not whether we treat him as an end in himself. If we train him to do the right thing by giving him treats, we view him as a means to our ends. That just makes him happy, and why not?
If the sole moral issue with animals is their happiness, we are in the realm that philosophers call “consequentialist,” where you add up all the benefits and subtract the harms. You don’t worry as much about bright lines. For example, eating less meat may enhance animal welfare if it reduces financial support for factory farming. But zero pounds of meat is just a number, like any other. Reducing your consumption from 50 lbs to 40lbs is ten times more important than getting it down from 1lb to zero. The same is not true with eating human flesh, which we regard as a matter of transgression and pollution. Even if cannibalism is merely a taboo, killing other people is truly wrong, and you’re a killer even if you only have one victim. I don’t think that’s the case with animals.
As long as our reasoning is consequentialist, offsets seem appropriate. It could be much better to eat a steak and contribute to an animal-welfare organization than to shun the meat but do nothing about public policy. Offsets and compensatory payments do not excuse violations of human rights, but they make sense with respect to animals (and nature more generally).