We Have No Idea If There Are Cost-Effective Interventions Into Wild-Animal Suffering

Cross-posted from Ozy Brennan’s personal website with minor edits.

I sometimes see people claiming very confidently that wild-animal welfare is completely intractable and there are no cost-effective interventions we can do to improve wild animals’ welfare. (Exact implications of this claim generally depend on the speakers’ values.)

This is honestly a quite extraordinarily claim. Think of all the ways human beings affect wild animals already: bird feeders, wildlife tourism, hunting, pest control, failing to adequately secure our dumpsters, disease control, air and water pollution, climate change, outdoor pets, invasive species, and so on and so forth. Are you telling me that there is not literally one of the dozens of ways we affect wild animals that has a knowable positive or negative effect on them? Perhaps some deity with a particularly odd ethical system has cursed us to an eternal neutrality, such that every rat we kill humanely will cause another rat to die horribly of poison?

It is not difficult to find a counterexample to this claim. Wildlife tourism is bad for wildlife unless it is particularly carefully done; it is also often bad from a conservation perspective, although of course that depends on the counterfactual, since wildlife tourism is no doubt better for biodiversity than the land becoming a freeway. (If you wish to engage in wildlife tourism, try to find proprietors that follow responsible tourism guidelines.)

Sometimes “wild-animal welfare is completely intractable” is used in a specialized sense, to mean “I accept certain philosophical arguments that mean that wild-animal lives are not worth living, and I don’t think there’s anything humans can do to cause their lives to be worth living.” However, if you don’t think we should wipe out nature from the earth (which, in my experience, most people who accept those arguments don’t), it is still possible to make things better or worse for wild animals, and (if cost-effective) it may be desirable to do so even if wild-animal lives aren’t worth living.

It may also be used to mean “there are not any wild-animal suffering interventions that are comparable in cost-effectiveness to GiveWell top charities.” I would not be surprised if this were the case. But I’m not sure how anybody could know that.

This is not, to be clear, because ecosystems are somehow inherently unknowable. It is true that ecosystems are very complicated and anything you do can have a dozen knock-on effects you never predicted. But we do, in fact, reason about what actions to take about ecosystems, even given our great uncertainty. Many people solemnly say that it is impossible, simply impossible, to know the effects of any action on ecosystems and therefore it would be irresponsible to take any action to protect wild animals– and then they eat wild-caught fish. Or let their cat go outside. Or put up a bird feeder. Or donate money to the Nature Conservancy. If ecosystems are unpredictable how do you know that nature preserves are a good way of preserving biodiversity anyway? Maybe things would be even more biodiverse if we cut down every tree in the rainforest!

The answer to this claim is that while of course ecosystems are dynamic and unpredictable systems and it is impossible to state with literally 100% certainty that cutting down the entire rainforest would be a bad conservation strategy, we do possess things like “nonzero level of knowledge about ecology” and “common fucking sense” that point to destroying their habitat being a poor way of protecting endangered species. Similarly, while there are tragic and costly mistakes, we can mostly figure out optimum sustainable yields for fisheries; most of the problem is in getting people to follow them instead of fishing as much as they damn well please. It is possible to know things about complicated systems with sufficient certainty that action is a better idea than nonaction.

The problem is that the wild-animal welfare space includes maybe a dozen people, nearly all of whom are dividing their time between wild-animal suffering and something else. As far as I know exactly one of us is a biologist; most people who do research about wild animal suffering are, by training, philosophers, social scientists, or programmers.

To be clear, this is a really terrible state of affairs. I as much as anyone want wild-animal suffering research to be done by people who have any discernible expertise in the field whatsoever. In my ideal world the field would consist of conservation biologists, wildlife managers, ecologists, ethologists and other people who can apply their academic knowledge to the question of improving wild animal welfare. However, this is somewhat difficult, because (a) only a few thousand people have heard of the concept of caring about wild animals’ welfare at all and (b) very very few scientists want to work part-time for minimum wage.

GiveWell benefited from a lot of development economics and public health research; they had to synthesize the fields, which is– to be clear– very important and very complicated, but once they did they could state their conclusions with a good deal of rigor. Animal Charity Evaluators is on shakier ground because academics tend not to find “how do we best make people vegan?” an interesting question, but at least they could benefit from decades of research about factory-farm conditions. Wild-animal welfare is a completely new field. The knowledge we need often exists– scattered across epidemiology, wildlife management, ecology, and a dozen other fields– but no one has ever collected it and applied it seriously to the issue of wild-animal welfare. It would take years simply to collect what is currently known, much less do any original research or begin to make intervention recommendations with reasonable cost-effectiveness numbers attached.

For most possible interventions– disease control, predator control, wildlife contraception, supplemental feeding, and so on–we don’t even know whether doing the thing would be good or bad. Again, not because it’s unknowable; just because there’s a limited amount you can do with twelve nonexperts working part time.

It is very easy to slide from “we do not know this” to “this is in principle unknowable and it is a waste of time to research it.” But resist the urge. Just because we don’t know something right now is no reason not to spend time trying to figure it out.