Farmed animals may have positive lives now or in a few decades?

[Vasco Grilo🔸]

The views expressed here are my own, not those of my employers or reviewers of the draft.

Summary

• I investigate whether farmed animals may have positive lives now or in a few decades. My results suggest the following:

  • Have negative lives. Hens in conventional cages, broilers in a conventional scenario, and decapod shrimp on an ongrowing farm.

  • May have negative or positive lives. Hens in cage-free aviaries, and broilers in a reformed scenario.

• I would say at least chickens’ lives can become positive over the next few decades in some animal-friendly countries.

• I am quite uncertain about the time when farmed animals of a given species will have positive lives in a certain region, if ever. To minimise the risk of decreasing the welfare of farmed animals, I think one should prioritise:

  • Improving the living conditions of farmed animals over decreasing the number of farmed animals with supposedly negative lives.

  • Learning more about:

    • The welfare of farmed animals by species and region.

    • The timeline of the effect of interventions aiming to decrease the number of farmed animals.

Introduction

In my mind, decreasing the number of farmed animals is only good if it increases welfare. This could be achieved by:

• Making the lives of farmed animals less negative or more positive.

• Decreasing the number of farmed animals with negative lives.

• Increasing the number of farmed animals with positive lives.

I recently updated my guess for the intensity of disabling pain to a value 10 % as high as before, and Julian Jamison had noted that relatively small changes to some previous guesses of mine for pain intensities could make farmed animals have positive instead of negative lives, which I agreed with. In this post, I investigate whether farmed animals may have positive lives now or over the next few decades.

Relatedly, see:

• Moritz Stumpe’s Lives not worth living?.

• Christoph Hartmann’s Are Organically Farmed Animals Already Living a Net-Positive Life?.

Methods

Overview

I estimate the welfare in animal quality-adjusted life years (AQALYs) adding that from pain and pleasure.

I compute the welfare from pleasure from the product between:

• The lifetime minus 8 h/​d of null welfare minus the sum of the time in hurtful, disabling and excruciating pain.

• The intensity of hurtful pain.

I calculate the (negative) welfare from pain from the negative of the sum of the contributions of the 4 categories of pain defined by the Welfare Footprint Project (WFP), annoying, hurtful, disabling and excruciating pain. I determine each of the contributions from the product between:

• The intensity of the pain as a fraction of that of a fully healthy life.

• Time in the pain in years.

I also express the welfare from pain as a fraction of the lifetime. The welfare as a fraction of the lifetime would be 1 AQALY/​year for the practically maximally happy life^[1].

Pain intensity

I rely on 2 sets of pain intensities:

• My guesses that:

  • Annoying pain is 10 % as intense as fully healthy life, such that 10 days (= ¹⁄₀.1) of annoying pain neutralise 1 day of healthy life.

  • Hurtful pain is as intense as fully healthy life.

  • Disabling pain is 10 times as intense as fully healthy life.

  • Excruciating pain is 100 k times as intense as fully healthy life.

• Means of lognormal distributions with 5th and 95th percentiles equal to the lower and upper bounds of Laura Duffy’s guesses^[2]. According to the means:

  • Annoying pain is 1.45 % as intense as fully healthy life (5th to 95th percentile, 1 % to 2 %).

  • Hurtful pain is 16.4 % as intense as fully healthy life (10 % to 25 %).

  • Disabling pain is 5.04 times as intense as fully healthy life (2 to 10).

  • Excruciating pain is 98.6 times as intense as fully healthy life (60 to 150).

The assumptions for the pain intensities imply each of the following individually neutralise 1 day of fully healthy life:

• For my assumptions:

  • 10 days of annoying pain.

  • 1 day of hurtful pain.

  • 2.40 h (= ²⁴⁄₁₀) of disabling pain.

  • 0.864 s (= 24*60^2/​(100*10^3)) of excruciating pain.

• For Laura’s assumptions:

  • 69.0 days (= ¹⁄₀.0145) of annoying pain.

  • 6.10 days (= ¹⁄₀.164) of hurtful pain.

  • 4.76 h (= ²⁴⁄₅.04) of disabling pain.

  • 14.6 min (= 24*60/​98.6) of excruciating pain.

Time in pain and lifetime

For the time in pain and lifetime:

• For broilers in a conventional or reformed scenario, I use data from WFP.

• For hens in a conventional cage or cage-free aviary, I use data from WFP.

• For decapod shrimp on an ongrowing farm with air asphyxiation, ice slurry or electrical stunning slaughter, I use data from Rethink Priorities (RP), and some assumptions for the 1st and 3rd of those slaughter methods.

Results

My pain intensities

Animal	Hen in a conventional cage	Hen in a cage-free aviary	Broiler in a conventional scenario	Broiler in a reformed scenario	Shrimp on an ongrowing farm with air asphyxiation slaughter	Shrimp on an ongrowing farm with ice slurry slaughter	Shrimp on an ongrowing farm with electrical stunning slaughter
Welfare for my pain intensities (AQALY)	-1.22	-0.178	-0.157	6.62*10^-4	-2.76	-1.39	-1.32
Welfare for my pain intensities as a fraction of the lifetime (AQALY/​year)	-0.908	-0.133	-1.29	0.00432	-8.77	-4.40	-4.19
Welfare as a fraction of the lifetime relative to that of the worst conditions	0	85.3 %	0	100 %	0	49.8 %	52.3 %

Laura’s pain intensities

Animal	Hen in a conventional cage	Hen in a cage-free aviary	Broiler in a conventional scenario	Broiler in a reformed scenario	Shrimp on an ongrowing farm with air asphyxiation slaughter	Shrimp on an ongrowing farm with ice slurry slaughter	Shrimp on an ongrowing farm with electrical stunning slaughter
Welfare for Laura’s pain intensities (AQALY)	-0.272	-0.0149	-0.0296	-0.00363	0.0112	0.0129	0.0130
Welfare for Laura’s pain intensities as a fraction of the lifetime (AQALY/​year)	-0.203	-0.0111	-0.243	-0.0237	0.0357	0.0409	0.0412
Welfare as a fraction of the lifetime relative to that of the worst conditions	0	94.5 %	0	90.3 %	0	14.6 %	15.3 %

Discussion

Farmed animals may have positive lives

Now?

My results suggest the following:

• Have negative lives. Hens in conventional cages, broilers in a conventional scenario, and decapod shrimp on an ongrowing farm.

• May have negative or positive lives. Hens in cage-free aviaries, and broilers in a reformed scenario.

I have not accounted for other animals linked to the ones I analysed, such as the male chicks linked to hens, and breeders linked to hens and broilers. Nevertheless, I do not know whether the other animals have higher/​lower welfare per lifetime than the ones I analysed, so it is unclear to me whether they make it easier/​harder to reach neutrality. In any case, I do not expect the other animals to change the overall picture that much. For:

• Hens:

  • I guess the welfare from pain per male chick life is less than 10 % that per hen life.

• Broilers:

  • I calculate the number broiler breeders killed per year in the European Union (EU) is 0.800 %^[3] (= 60*10^6/​(7.5*10^9)) of the number of broilers for consumption killed there per year.

  • Broilers in a conventional scenario “in the EU and the US” live for “42 and 47 days”, i.e. roughly 6.36 weeks (= (42 + 47)/​2/​7). Broiler breeders live “55–62 weeks”, i.e. around 58.5 weeks (= (55 + 62)/​2), or 9.20 (= 58.5/​6.36) times as long as broilers in a conventional scenario.

  • From the 2 points above, I infer there are 7.36 % (= 0.00800*9.20) as many broiler breeders as broilers for consumption.

  • Supposing the welfare from pain per lifetime is 2 times as high for broiler breeders as for broilers for consumption, which I take to be pessimistic because breeders live longer, the welfare from pain linked to broilers would increase by 14.7 % (= 0.0736*2).

Over the next few decades?

Even if one is certain a given population has negative lives now, decreasing it could still be bad if it prevents the existence of positive lives in the future. I believe this possibility is often overlooked. For example:

• I did not discuss it when I listed a bunch of effects of decreasing the number of farmed animals.

  • The closest I got was saying I did not know whether some populations of animals have positive/​negative lives, thus being uncertain about the value of decreasing their consumption.

• James Özden did not cover it in the posts Theories of Change for the animal advocacy movement and The default trajectory for animal welfare means vastly more suffering.

  • These seemingly assume that decreasing factory-farming is necessarily good, which may not hold if the lives of farmed animals become positive in the future.

  • I think the posts are great anyways! They are among my favourite high-level posts about animal welfare.

There would be no concern if one targeted decreasing populations of farmed animals whose lives are and will for the next few decades continue to be robustly negative. However, I believe this is unclear even now for chickens in improved conditions. It is also worth having in mind that Open Philanthropy, the main funder of both cage-free and broiler welfare campaigns, only started supporting these in 2016, i.e. just 8 years (= 2024 − 2016) ago. So I would say at least chickens’ lives can become positive over the next few decades in some animal-friendly countries, like ones in the EU, the United Kingdom, and the United States. The same may apply to other species^[4].

As a side note, I suppose the slower egg production of hens in cage-free aviaries, and slower growth of broilers in a reformed scenario may end up being doubly beneficial if they eventually transition to systems where they have positive lives. In this case, besides resulting in higher welfare per chicken-year, they would lead to a larger population of chickens with positive lives. In contrast, if hens and broilers still have negative lives in the improved conditions, one has to ensure the increase in welfare per chicken-year is larger than the increase in population for the change to be beneficial in the nearterm.

Consequences given uncertainty

I am quite uncertain about the time when farmed animals of a given species will have positive lives in a certain region, if ever. There is also huge uncertainty about which wild animals have positive/​negative lives, and how their population sizes are affected by changes in the number of farmed animals. To minimise the risk of decreasing the welfare of farmed animals, I think one should prioritise:

• Improving the living conditions of farmed animals over decreasing the number of farmed animals with supposedly negative lives.

  • For instance, supporting corporate campaigns and welfare laws/​standards over dietary change.

  • If one insists in supporting interventions aiming to decrease the number of farmed animals with supposedly negative lives, I believe ones targeting nearterm decreases are preferable, because I guess farmed animals are more likely to have positive lives in the future. I am not confident investments in alternative proteins are beneficial given their long time horizons.

• Learning more about:

  • The welfare of farmed animals by species and region, expanding the research of WFP by determining the time spent in the 4 categories of pleasure they defined (satisfaction, joy, euphoria and bliss), and investigating pain and pleasure intensities (relatedly).

  • The timeline of the effect of interventions aiming to decrease the number of farmed animals.

I strongly endorse expected total hedonistic utilitarianism, but I imagine the above conclusions are reinforced by moral uncertainty. Improving the conditions of animals necessarily increases both total and per capita animal welfare (holding the number of animals constant), whereas decreasing the number of animals with negative lives only necessarily increases total animal welfare (holding the conditions constant), and there are moral theories which care about welfare per capita (holding total welfare constant).

Acknowledgements

Thanks to Derek Shiller for feedback on the draft.

1. ^

   The welfare per time of the practically maximally happy life is much lower than that of the maximally happy instant.

2. ^

   Laura was executive research coordinator at Rethink Priorities from June 2022 to November 2023.

3. ^

   “Annually about 7,500 million broilers are reared in the EU27. To produce these broilers about 60 million broiler breeders are required.”

4. ^

   It is hard to tell, as I am not aware of data on the time spent in pain by species besides chickens in improved conditions. For reference, WFP is working on projects on fish and pigs.

[Alex Schwalb]

Very interesting article! Although I would disagree that it would be bad to decrease the number of factory farmed animals if they have positive lives. What we’re doing when decreasing the number of factory farmed animals is just shifting the biomass to be in different forms. I think humans are capable of much more positive lives than farmed animals, so in the long term future it would be best to have as much biomass in the form of humans (and possibly pets) as possible. A world where humans eat predominantly plants and cultivated meat would be able to support more humans, and these extra humans would have much better lives than farmed animals.

When it comes to shifting the biomass towards wild animals, I don’t know whether it would be good or bad though. I think in the long term future after people start intervening, wild animals would probably also have better lives than farmed animals, because people would value them intrinsically instead of instrumentally. Farms will always be optimised to produce as much output as possible, whereas future “nature reserves” could be optimised for welfare

[Vasco Grilo🔸]

Thanks for the comment, Alex! I strongly upvoted it because I like that you tried to think about how to increase welfare assuming farmed animal end up with positive lives, instead of dismissing this as impossible, or arguing that factory-farming is intrinsically bad.

I think humans are capable of much more positive lives than farmed animals, so in the long term future it would be best to have as much biomass in the form of humans (and possibly pets) as possible. A world where humans eat predominantly plants and cultivated meat would be able to support more humans, and these extra humans would have much better lives than farmed animals.

I agree humans are capable of more positive experiences that animals, but not that much more. I also agree plant-based foods would enable supporting more humans. However, to maximise welfare, one should look for interventions which increase welfare the most per $. At least now, I think these are ones helping animals, not humans (i.e. not the species whose individuals are capable of experiecing the most welfare). I estimate:

• Broiler welfare and cage-free campaigns (helping chickens) are 168 and 462 times as cost-effective as GiveWell’s top charities.

• Shrimp Welfare Project’s Humane Slaughter Initiative is 64.3 k times as cost-effective as GiveWell’s top charities.

I expect helping animals will continue to be more cost-effective than helping humans longerterm, at least given humans’ current form, because animals have a higher ratio between welfare range and calorie consumption^[1].

Species	5th percentile welfare range per calorie consumption as a fraction of that of humans	Median welfare range per calorie consumption as a fraction of that of humans	95th percentile welfare range per calorie consumption as a fraction of that of humans
Bees	0	4.88 k	31.7 k
Shrimp	0	83.9	3.11 k
Crayfish	0	17.5	226
Salmon	0	3.61	33.1
Chickens	1.50 %	2.49	6.52
Humans	1.00	1.00	1.00
Pigs	0.459 %	47.3 %	94.7 %

1. ^

   The welfare range is the difference between the welfare per time of a practically maximally happy and unhappy life.

[Alex Schwalb]

Thanks for the reply. I completely agree that we should look for interventions that improve welfare most per $, and that those, at least for now, are the ones focusing on animals and not humans. 100% of my donations at the moment actually goes to animal causes.

That’s a very interesting table about welfare range per calorie consumption. It caused me to update away from my belief that in the ideal far future we should dedicate most resources to creating more happier humans (or the next generation of the most sentient beings), and towards the belief that the existence of some lower sentience but super efficient beings, such as bees, would be a great thing. It doesn’t change any near term things for me or have any actionable consequences though, as I think we’re still gonna be in the “reduce suffering” part of history for a very long time before we get to the “increase bliss” part

[alene]

This is a really interesting article. Thank you for writing it. I hope it’s true that farmed animal welfare will one day be net positive. I fear that the treatment of farmed animals seems to be getting worse, rather than better, over the course of human history. But I hope I’m wrong. Or maybe it’ll be a boomerang-shaped change, where the treatment of animals is currently getting worse, but things are about to do a 180 and start moving in the other direction? I hope we EAs can make that happen.

[Vasco Grilo🔸]

Thanks, Alene!

I think total animal welfare is currently decreasing because factory-farming is increasing in low and middle income countries due to population and economic growth. However, I believe animal welfare per person is already increasing in countries where welfare reforms have been successful, and I expect these to eventually spread globally. Air pollution and greenhouse gas emissions also dramatically increased since the industrial revolution, started by going down in high income countries, and I expect they will eventually go down globally.

In any case, even if positive factory-farming was inevitable (some time in the future), getting there faster would still be super valuable!

[CB🔸]

Very interesting, thanks a lot !

Do you have any data on pigs or cows ? I know they represent a smaller number of animals, and there is less data from the welfare footprint project, but I’d be curious to know your opinion.

The calculations would also be complicated by the necessary inclusion of veal for cows, and gestation crates for pigs, since these are a necessary part of current systems and certainly bad most of the time.

But since grass-fed cows are likely to be better treated than other animals, it would be good to know if their lives are positive and can maybe provide a template for other forms of animal agriculture (although economic incentives don’t push in this direction).

[Vasco Grilo🔸]

Thanks, CB!

Do you have any data on pigs or cows ?

WFP is working on a project on pigs, and were hiring for a project on beef cows, but they only have data on chickens. I think the conditions of pigs and cows are better than those of hens in conventional cages and broilers in a conventional scenario, and I guess they are also better than those of hens in cage-free aviaries and broilers in a reformed scenario, but not by a lot. I also guess cows have better conditions than pigs. In sum, I would say:

• For my guesses for pain intensities, pigs have negative lives, but cows may have positive/​negative lives.

• For Laura’s guesses for pain intensities, cows and pigs may have positive/​negative lives.

The calculations would also be complicated by the necessary inclusion of veal for cows, and gestation crates for pigs, since these are a necessary part of current systems and certainly bad most of the time.

Nice point. As for male chicks and broiler breeders, I do not think accounting for veals changes the overall picture that much:

• According to Compassion in World Farming (CIWF), “Around six million calves are reared for veal within the EU every year”. Veals are defined in the EU as having less than 1 year. If veals are slaughtered at 0.5 years (= ¹⁄₂), their population is 3 M (= 6*10^6*0.5), i.e. 4.18 % (= 3/​(74.8 − 3)) of the number of non-veal cows in the EU.

• If non-veal cows have negative lives, and veal cows’ conditions are 2.5 times as bad, accounding for veals only decreases the overall welfare by 10.4 % (= 0.0418*2.5).

However, it looks like accounting for pigs in gestation crates may be more important:

• From Wikipedia, “There were 5.36 million breeding sows in the United States as of 2016, out of a total of 50.1 million pigs”, i.e. 12.0 % (= 5.36/​(50.1 − 5.36)) as many pigs in gestation crates as outside these.

• If pigs outside gestation crates have negative lives, and the conditions of pigs in gestation crates are 2.5 times as bad as those of pigs outside them, accounting for pigs in gestation crates decreases the overall welfare by 30.0 % (= 0.120*2.5).

But since grass-fed cows are likely to be better treated than other animals, it would be good to know if their lives are positive and can maybe provide a template for other forms of animal agriculture (although economic incentives don’t push in this direction).

I was referring to non-grass-fed cows above. For both my and Laura’s guesses for pain intensities, I think grass-fed cows have positive lives.

[CB🔸]

Thanks for the answer !

This is really interesting. Do you think that the fact cows are separated from their child, and arguably really don’t like that, would change significantly the results?

[Vasco Grilo🔸]

Thanks for another relevant question too! I do not think that alone would make dairy production net negative:

• According to CIWF, “Dairy cows must give birth to one calf per year in order to continue producing milk”.

• From Animal Australia, “both mother and calf can often be heard calling out for each other for hours [after they are separated]”.

• Assuming disabling pain of 3 h/​year for each the mother and child based on the above, one gets 6 h/​year (= 2*3) of disabling pain. For my intensity of disabling pain, that corresponds to a loss of 0.00684 AQALY/​year (= 6/​24/​365.25*10).

• The above is quite small in comparison with the magnitude of the values I got for chickens and shrimp. So, in the absence of longer term effects from the separation, I do not think it would bring dairy production from positive to negative.

[CB🔸]

Thanks for the answer ! I wish more people thought about these questions.