Animal Agriculture and the Non-Identity Problem

by Michael Dorf

On Saturday, Prof. Colb and I will be speaking at an animal rights conference at Bucknell University with a focus on approaches to the topic that aim to abolish animal agriculture. Our topic is "Who Benefits From Abolition?" It is framed as a response to a particular sort of objection to the case for veganism. Consumers of animal products sometimes say that their acts are harmless or even beneficial to the animals whose parts and products they consume, because those animals otherwise wouldn’t have existed. This claim appears to raise what philosophers call the non-identity problem.

As philosopher David Boonin explains in his fascinating 2014 book, the non-identity argument reasons from a number of seemingly correct premises--such as the premise that an act that harms no one is not wrong--to an implausible conclusion: that an act that is both necessary to a person's existence and leads to serious challenges for that person is not wrong, so long as that person has a life worth living. This tension is what makes the non-identity problem a problem.

A number of years ago, in discussing the non-identity problem, I gave the example of a drug that both boosts fertility and causes birth defects. Can the manufacturer of the drug be said to have wronged the children born with birth defects as a result of their mothers having taken the drug? Absent the drug, the children not only wouldn't have the birth defects; they wouldn't exist, because their mothers would have been infertile. If, despite the birth defects, the children have lives worth living, then how have they been wronged?

I have since learned that in the traditional formulation of the non-identity problem, it is necessary that the person making the relevant decision face a choice between (at least) two options, one of which results in the person with the disadvantage and the other of which results in a different person without the disadvantage. To my mind, that feature of the problem is not crucial to the moral question, but I'm willing to go along with the philosophers, so we can use a different example.

For instance, Boonin hypothesizes a couple who have a choice between conceiving a child immediately but the child will have some serious disability that nonetheless results in a life worth living, versus taking a pill with no adverse side effects for a few weeks and then conceiving a different child (because from a different egg and presumably a different sperm) without the disability. This then poses the nonidentity problem if the couple do not wait: The resulting child has a disability but cannot say that he was wronged, because the act that would have avoided the disability--his mother taking the pill for a few weeks while delaying conception--would also have prevented his existence.

Even accepting the choice requirement, the non-identity problem remains relevant to the animal exploitation example. Consuming animal products fuels demands for those products, which results in land (and in the case of farmed fish, water) being used for the raising of the animals for food and fiber, which in turn leaves fewer resources for wild animals.

Most of the literature regarding the non-identity problem aims to show that one of the premises that goes into the non-identity argument is wrong. In contrast, Boonin contends that none of these approaches succeeds and argues instead that we ought to accept the implausible conclusion as not that implausible after all. He thinks the non-identity argument succeeds.

In our remarks on Saturday, we intend to offer three sorts of responses to this argument. One response is that most of the animals raised for food and fiber do not have lives worth living, because the conditions in which they are raised are miserable. The short life of these animals is a curse, not a blessing. We think that's true even for many of the supposedly "humanely raised" animals.

Suppose, however, that some animals raised for food or fiber do, all things considered, have lives worth living. Even for these animals, it is not clear that the non-identity argument works. If Boonin is wrong--that is, if the implausible conclusion really is less plausible than one or more of the challenges to the premises of the non-identity argument--then whatever reasoning dissolves the non-identity problem in other contexts will dissolve it for animals as well. We will explore a variant of one of the arguments Boonin rejects in our remarks tomorrow, but we don't want to set it out here both for reasons of brevity and because our view is still tentative.

We are even more tentative about the third response. Indeed, at this point I haven't fully persuaded Prof. Colb of this point, so in the balance of this post I speak only for myself. This final point aims to show that despite appearances, the argument that animals raised for food and fiber wouldn't exist but for demand for food and fiber is mostly not a non-identity argument at all.

I want to say that even if Boonin is right in his bullet biting with respect to the genuine non-identity problem, that problem is not presented by the animal exploitation argument. I would emphasize a key difference from the classic versions of the non-identity problem. In the classic versions, the misfortune that befalls the person as a result of the act that is also necessary for her existence is, as it were, baked into the state of the world from the moment she comes into existence. She is born with the disability. Or, in another classic example, she is born into a world that is seriously polluted (but wouldn't have been born, due to the precariousness of any individual's birth given butterfly effects, in a less polluted world).

To be sure, some of the harm that befalls animals raised for food and fiber is baked into their existence. For example, domesticated hens bred to produce many more times as many eggs as the wild birds from which they descended will predictably suffer calcium deficiencies and uterine prolapses. Turkeys and pigs bred to grow very big very fast suffer from their size, even if they are quickly rescued from the food industry and moved to sanctuaries.

But much of the harm that befalls animals raised for food and fiber--even for the small fraction whose lives are worth living--comes as a result of actions taken after they come into existence and is specifically associated with exploiting them. The mother cows have their calves taken away so that the milk can be diverted to human consumption. The animals raised for meat who are, by hypothesis, leading lives worth living, have those lives cut short by slaughter.

We cannot say of these animals that but for the act that harms them, they would not exist, because they must first be brought into existence and only then subject to the harm. It's true that but for the plan to subject them to harm, the farmer never would have brought them into existence, and thus, in a sense, the consumer who demands animal products can take some "credit" for the animals' existence in virtue of the exploitation of those animals. However, this sort of argument proves too much. People brought into existence to be exploited--for their organs, say, as in Kazuo Ishiguro's novel Never Let Me Go, or for their slave labor in societies that practice slavery--would not exist but for the intention to subject them to exploitation, but we nonetheless properly regard the exploitation of them as wrongful once they exist. The non-identity literature includes considerable discussion of the question whether descendants of enslaved persons who would not exist but for slavery were wronged by slavery, but there can be no doubt that the people who were themselves enslaved were wronged by slavery. Once they were born into slavery, their "owners" had a moral duty to free them, even if their lives as slaves were worth living and even if they would not have been born but for slavery, because continued enslavement was a continuing wrong. Organ farming and enslavement of humans are not instances of the non-identity problem, and therefore, neither is the harm done to non-human animals in exploiting them defensible by the invocation of the non-identity problem.

Finally, it might be thought that virtue ethics provides another way around the non-identity problem. When I've described the problem to friends and colleagues, some have said that even if the choice to do the thing that results in the disadvantage relative to the other beings doesn't wrong the beings thereby called into existence, it is somehow wrong in itself. I have some sympathy for this view, just as I have some sympathy for at least some of the claims of virtue ethics more broadly. It does seem to exhibit bad character to make the choice that brings about a worse state of the world, even if it is not a state of the world that is worse for any of the individuals in that state of the world, who wouldn't exist in any other state of the world. But I don't think that virtue ethics fully captures what's wrong with the non-identity argument, because the misfortune that befalls the (human or non-human) being  in the worse state of the world feels like a harm to that being. The second argument to which I adverted but did not describe above aims to capture how that intuition might be rendered plausible.