Rewilding Is Extremely Bad

Crosspost of my blog article. 

Rewilding is the process of converting fairly substantial tracts of industrial land back into nature. What once was farmland or a logging area returns to its natural state. The consensus among most people tends to be that it’s a pretty uncontroversially good idea. This consensus, I think, is badly wrong. Rewilding is extremely immoral—worse than almost any other thing we do—and we should refrain from it barring exceptional circumstances.

What’s so bad about rewilding? The basic case against is that it majorly increases wild animal suffering. If we assume wild animals are spread uniformly across Earth’s land, then each square mile of land contains 1,754-17,540 mammals, 1,754-1,754,000 reptiles, and 1,754-1,754,000 amphibians.

Where things get really extreme is with insects.

On average, a square foot of land contains about 750 insects. This means that if you rewild a square mile of land, over the course of a year, 20.9 billion additional insect life-years will be lived. If we assume that each insect lives two weeks—a fairly reasonable estimate—then a mile of rewilded land produces about 585,480,000,000 (five-hundred-eighty-five-billion-four-hundred-eighty-million) extra insect lives and deaths annually.

Let’s assume conservatively that half that many insects would have lived and died on the land if it wasn’t rewilded and that insects suffer 1% as intensely as people (which I think is an underestimate in expectation). Then we get the result that rewilding a single square mile of land produces about as much pain as that of nearly 2,927,400,000 extra human deaths. And if we count soil animals, then things get worse, in expectation, by orders of magnitude. Each square mile of land rewilded is a very serious moral atrocity. There could very well be more suffering in a thousand miles of rewilded land over the course of a few years than all suffering that’s existed so far in human history.

Now, it would be one thing if this suffering was compensated by a similar increase in well-being. But it’s not. As I’ve argued at length, nature contains vastly more suffering than well-being. Most animals live only a few days or weeks and die painfully. They have nowhere near enough well-being over the course of their life to outweigh their extremely painful deaths.

If you lived just a week, and then starved to death or got eaten alive, your life wouldn’t have been worth living. But this is the fate of almost all animals who have ever lived. Nearly all of them have very short lives that culminate in a painful death. A few fleeting moments of joy aren’t enough to compensate for the intense suffering at their end of life.

And while there might be deontological prohibitions against destroying nature, these certainly don’t justify bringing nature back. There is no reason to bring nature back, forcing countless beings to live a brief life of misery, terror, and agony. Compassion demands better. If there are deontic prohibitions in this area, then they certainly militate against forcing countless beings to be eaten alive and starve to death.

Now, what defenses are there of rewilding? One is that nature is very pretty. Humans benefit from having nature nearby. These benefits are small and uncertain. They’re completely swamped by the impact on wild animals.

Others suggest that nature is a good thing intrinsically. I find this perspective baffling; if nearly every creature in nature is running for its life, very soon to be dead, likely to starve or be eaten alive by the end of the week, then nature isn’t worth preserving. It’s extremely implausible that nature has intrinsic value—if you try to make the view precise, it collapses completely. More importantly, even if nature has some intrinsic value, it doesn’t have enough to justify causing more suffering than has ever existed in human history, every few years, in relatively small plots of land.

Now, maybe you just don’t care about the interests of wild animals. But this is a fairly baffling moral perspective—one that cares deeply about nature, but doesn’t care one lick for the beings in nature. If we should care about an abstraction like nature, then we should care about the real sentient beings that live in nature. An abstraction like nature never feels anything. An individual deer, ripped limb from limb by lions, does. This view strikes me as like caring about a country but not caring about any of the people in the country.

And it is very difficult to see why it is that beings in nature wouldn’t matter. Surely humans mattered a few hundred thousand years ago, even when we were part of nature. So merely being part of the natural ecosystem doesn’t strip one of moral worth. But then it’s hard to see what about wild animals would make them not deserving of moral consideration. One can’t say that it’s being part of nature and being unintelligent (as wild animals are) because that would imply that mentally disabled people didn’t matter before the dawn of civilization (when they were part of hunter-gatherer societies).

Come to think of it, it’s not at all clear that there’s a precise definition of being part of nature—arguments against caring about wild animals, then, inevitably turn into arguments against caring about animals at all. They’d then imply it would be fine to torture dogs for trivial reasons.

And as I’ve argued many times, it’s bad to suffer because of how suffering feels, not extrinsic characteristics like whether one is a wild animal. The fact that wild animals experience many thousands of times more suffering than all humans combined, in expectation, makes their suffering an urgent priority.

Now, the only serious argument for rewilding (if one grants that wild animals mostly live bad lives, which I think there is a very powerful case for) is that doing so can reduce existential threats. The argument goes: rewilding reduces existential risks by improving environmental quality and reducing global warming. Because the future could be so enormous, this outweighs the harms to wild animals in the present.

I find much about this argument extremely dubious.

1. The people who make it generally would not accept this logic in other domains. Virtually no one would support torturing billions of people if doing so would reduce existential threats by .00000000000000001%. We don’t normally think it’s worth causing hideous, unfathomable quantities of suffering to maybe reduce existential risks by a tiny amount.

2. Rewilding has no serious impact on existential risk. The primary existential risks it could reduce are from biodiversity and climate change. But neither of these are major existential risks. John Halstead, after producing the most detailed report on climate existential risks done ever, guessed the odds of climate change existential risk were way below .1%—there’s just no coherent mechanism by which it would spell doom, the world has survived past periods of significant warming without facing catastrophic collapse, and the impact of climate change on food production will be largely insignificant. And rewilding has relatively limited impact on climate change.

    

   The scenario for climate change causing extinction generally involves refugee crises triggering a nuclear war. This is very implausible; the places where climate change is likely to cause extra wars are mostly in Africa, which has no great powers. The overall impact on wars is likely to be fairly negligible, as the studies Halstead reviews find, and there’s no plausible mechanism by which such conflicts would go nuclear. Even the IPCC found that there’s little evidence for climate change causing interstate wars—rather than civil wars—which are the only kind that could plausibly go nuclear.

    

   Biodiversity loss as an existential risk is even more unlikely—you can read Halstead’s report for more detail but many places have lost nearly all their biodiversity without facing any societal crisis. When you account for most species loss being on island nations, biodiversity isn’t even declining by that much. And even a collapse of biodiversity wouldn’t endanger food sources; pollinators only are responsible for about 5% of crop yields, and the rest of biodiversity contributes little to food production. There is, of course, no coherent way that rewilding a bit would have any effect on the kind of biodiversity crisis that would kill off the pollinators.

    

   When you ask those who think that biodiversity loss will cause extinction how it would do so, you get one of two explanations. The first is pseudoscience, like that the loss of biodiversity will lead to us running out of Oxygen. This is laughable as Oxygen cycling works on scales of millennia and the net effect from biodiversity loss on Oxygen is likely to be very minor—a textbook on the subject recounts “changes in atmospheric oxygen concentration, although measurable, will be trivial until coupled, global-scale geologic/ biological processes conspire to change them. This would likely take millions of years.”

    

   The other way of arguing involves vague innuendo about tipping points and the like. It involves tossing about lots of apocalyptic words (brink, tipping point, and so on), but few specific scenarios. These risks are not generally taken seriously by climate scientists, but instead promulgated by activists. In light of the poor track record of such claims, and the fact that humanity has survived fine despite exterminating most European forests (contra claims of existential tipping points), they should not be taken very seriously unless backed up by strong evidence, which they never are. And in any case, rewilding would have a minuscule impact on the global climate, so if we’re otherwise doomed, rewilding is extremely unlikely to save us.

3. Given that existential risks affected by rewilding are so minor, they could very well be outweighed by the opposite risks—the risks caused by rewilding. Rewilded areas could introduce new diseases that pose existential risks, for example, or the economic delay could result in America losing a technological arms race to China. Given how low the risks of extinction are in the other direction, these could be more significant.
4. In my view a much bigger risk from rewilding is that we’ll get in the habit of rewilding, and then spread nature across the universe. This could be the cause of many terraformed planets—and utterly incomprehensible quantities of suffering. By delivering successful rewilding, a pro-nature coalition might be built, which would spread incomprehensible amounts of suffering to the stars. This could be an S-risk—a risk of astronomical suffering, on the order of existential risks in moral horror. And while I expect most future people to be digital, the impact on spreading wild animal suffering to the stars seems more significant than rewilding’s trivial impacts on existential risks. The values we have today will influence what we do in the future; if they are environmentalist pro-rewilding values, the future may be hell for countless quintillion beings.

5. As Bostrom has argued, we might currently be in a simulation. If we are, then our actions potentially are mirrored across all simulated worlds. If so, then rewilding doesn’t produce astronomical suffering just in our universe, but in all universes. Thus, as Brian Tomasik has argued, the simulation argument weakens the case for caring almost exclusively about the future, which is the only sound basis for prioritizing the minuscule possible existential risk reduction over the enormous present suffering caused by rewilding. A tiny reduction in the risks of the 8 billion people alive today is paltry compared to the impact on the quintillions of suffering wild animals.

    

   It would be one thing if we were comparing major impacts on existential risks to major suffering reductions. But we’re not. We’re comparing tiny and highly uncertain in terms of their overall sign impacts on existential risks to very sizeable increases in suffering. In such a choice, given moral uncertainty, one should refrain from causing huge amounts of extreme suffering.

    

   Now, don’t get me wrong: I’m pretty Longtermist. I think the most important thing to do today is make the future go well. But I’m not so Longtermist that I’m willing to cause extreme suffering to countless beings for a wildly unclear—and certainly tiny—impact on existential risks.

All of this is to say that if you are a consequentialist, you should be very opposed to rewilding. It causes countless innocent beings to suffer horribly. Because nearly all wild animals live short lives of intense suffering, rewilding is a major moral catastrophe—just about the worst thing we could do collectively.