Biorisk is smaller in scale than farmed animal welfare
Introduction
In this post, I argue that the scale of farmed animal suffering far exceeds the expected mortality risk from engineered pandemics—by a factor of at least 100. I haven’t seen this explicit comparison made elsewhere, yet it strongly affects my personal cause prioritization.
However, it’s important not to update solely on scale: the tractability and neglectedness of pandemic preparedness may still justify prioritizing it, despite its seemingly smaller scale. I would be excited to see more detailed work in this area—for example, comparing the Disability-Adjusted Life Years (DALYs) lost to farming versus pandemics more rigorously than I do here.
Scale of biorisk
In The Precipice (2020), Toby Ord estimates a 1⁄30 risk of human extinction from engineered pandemics[1] this century. Spreading this risk over 80 years—from 2020 to 2100—gives an approximate 1⁄2,360 chance of extinction per year (on average).
The average global population in the 21st century may be around 9 billion people. Multiplying this by the annual extinction probability implies an expected human death toll of roughly 3.8 million per year (i.e., 9 billion × 1⁄2,360).
Of course, DALYs lost to pandemics would be higher than just deaths, since pandemic survivors can suffer severe morbidity and long-term health consequences. Further, this calculation is a simplification in that it treats the extinction risk as evenly distributed each year, which is not necessarily accurate. The true distribution of risks might be heavily skewed toward certain time periods. [1]
Scale of industrial farming
Number of Animals
According to Our World in Data, 83 billion land animals and 128 billion farmed fish were slaughtered globally in 2022 for meat alone. While this number has been rising, I will assume for simplicity that it remains constant for the rest of the century (on average).
Converting to human-equivalents
Rethink Priorities (RP) estimates that all land animals we farm have a welfare range bigger than 0.1, and all fish higher than 0.01. Using these figures, we might approximate each land animal as 0.1 of a “human-equivalent” in terms of potential suffering, and each fish as 0.01 of a human-equivalent. Multiplying:
83 billion land animals × 0.1 = 8.3 billion human-equivalents
128 billion fish × 0.01 = 1.28 billion human-equivalents
Summing these gives ~9.58 billion human-equivalents in suffering per year as a lower bound, compared to RP’s numbers. Note that this calculation assumes that the “badness”/disutility of an animal dying is proportional to its welfare range—a stance that RP doesn’t explicitly endorse. Here, I simply use the welfare range values as a rought proxy.
Caveats and Uncertainties
These welfare range estimates are inherently subjective and vary between researchers. My personal guess is somewhat lower than RP’s, but not by more than an order of magnitude. If your range estimates differ by more, we might get different conclusions.
The duration of suffering versus the duration of a human life is not directly accounted for here.
DALYs lost likely exceed these simple “human-equivalent” calculations, since many animals endure intensely negative conditions for most or all of their lives.
RP’s weights aren’t meant to compare lives in this fashion. Here, I use them more as a proxy than anything, which is part of why I round down significantly.
Why I might be wrong
Apples to Oranges
The figure from Toby Ord focuses on extinction-level pandemics. Large, non-extinction pandemics could happen more frequently, increasing expected annual human deaths beyond the 3.8 million estimate without increasing extinction risk. If such mid-level pandemics become common, total pandemic-related mortality might be much larger than implied here.
Population Ethics
Under a totalist view , a catastrophic pandemic that kills a fraction of humanity (without total extinction) might still massively reduce our long-term population, shrinking the number of future lives. The classic case for existential risk work often looks beyond literal extinction to consider any permanent curtailment of civilization’s potential.
I personally subscribe more to a person-affecting view, which means I place less weight on future people not getting the chance to exist compared to people already existing being negatively affected. If you’re not a longtermist, this point will also be less relevant.
Civilizational Collapse and Value Lock-In
A sufficiently severe pandemic might cause societal collapse, after which a new civilization can emerge with entrenched institutions that “lock in” certain values or systems. This lock-in could manifest as:
An authoritarian system that oppresses humans.
A continuation or scaling up of current factory-farming practices, resulting in the same or even greater animal suffering.
(Credit to Isaac Dunn for this idea.)
Animal Suffering from Pandemics
During the last pandemic, Denmark had to put down roughly 17 million minks over concerns of viral mutations. For bigger pandemics, we can expect similar or even more extreme measures affecting animals—especially if civilizational collapse occurs and less regulated “emergency” actions become commonplace. This possibility suggests that impacts on non-human animals from pandemics could be significant enough to warrant more biorisk research than the above comparison alone might indicate.
(Credit to Sam Smith for this idea.)
- ^
At the extreme, the risk could be 0 every year except for one year when it is 1⁄30. This would give an expected death toll of 333 million that year, which is equivalent to 4.17 million per year—only slightly bigger.
It’s pretty crucial how much less weight you place on future people, right? If you weight there lives at say 1/1000 saving the life of a current person, and there are in expectation going to be 1 million x more people in the future than exist currently, then most of the value of preventing extinction will still come from the fact that it allows future people to come into existence.
I buy that. One way of putting it would be to say that if you use a parliamentary method of resolving moral uncertainty, the “non-totalist population ethics rep” and the “non-longtermist rep” should both say that farmed animal welfare as greater in scale than biorisk. Does that seem more useful?
I don’t know enough about moral uncertainty and the parliamentary model to say.
It’s worth saying that although in EA, people favour approaches to moral uncertainty that reject “just pick the theory that you think is mostly likely to be true, and make decision based on it, ignoring others”, I think some philosophers actually have defended views along those lines: https://brian.weatherson.org/RRM.pdf
The motivation for focusing on global catastrophic risks is that these could dramatically limit humanity’s potential. If, per your population ethics, such a limitation wouldn’t be concerning, then it’s not surprising that you wouldn’t find work aiming to avert or mitigate such risks compelling.
I think the post would be clearer if it were explicit about this up front: the disagreement here isn’t about the relative scale of biorisk vs factory farming, but instead about how much value there is in averting civilizational collapse and/or extinction.
I’m sorry if the title was misleading, that was not my intention. I think you and I have different views on the average forum user’s population ethics. If I believed that more people reading this had a totalist (or similar) view, I would have been much more up front about my take not being valid for them. Believing the opposite, I put the conclusion you’d get from non-person-affecting views as a caveat instead.
That aside, I’d be happy to see the general discourse spell out more that population ethics is a crux for x-risks. I’ve only gotten—and probably at some points given—the impression that x-risks are similarly important to other cause areas under all population ethics. This runs the risk of baiting people into working on things they logically shouldn’t believe to be the most pressing problem.
On a personal note, I concede that extinction is much worse than 10 billion humans dying. This is however for non-quantitative reasons. Tegmark has said something along the lines of a universe without sapience being terribly boring, and that weighs quite heavily into my judgement of the disutility of extinction.
Executive summary: The scale of farmed animal suffering vastly exceeds the mortality risks from engineered pandemics, suggesting a need to prioritize animal welfare over biorisk mitigation.
Key points:
Farmed animal suffering is estimated to be at least 100 times greater than the mortality risk from engineered pandemics.
Annually, approximately 211 billion animals are slaughtered for meat, resulting in around 9.58 billion human-equivalents of suffering.
While biorisk focuses on extinction-level pandemics with an expected death toll of ~3.8 million per year, industrial farming’s impact on well-being is substantially larger.
Uncertainties include the subjective nature of welfare estimates, potential underestimation of DALYs from farming, and the possible severe consequences of pandemics leading to increased animal suffering.
The author highlights the need for more rigorous comparisons of Disability-Adjusted Life Years (DALYs) lost between farming and pandemics.
Recommendations include prioritizing animal welfare in cause prioritization due to its significantly larger scale of impact.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.