Faunalytics has releasedĀ a major update to our Animal Product Impact Scales (APIS), a resource that estimates the animal lives and days of suffering behind different animal products. The 2026 update refreshes the U.S. data and, for the first time, adds U.K. estimates. This update was led by Chris Bryant at Bryant Research.
If you think about animal welfare in terms of where marginal effort or funding goes furthest, APIS is designed to be directly useful. Hereās whatās new and why we think itās worth your attention.
What APIS Measures
APIS estimates two metrics for each animal product: animal lives taken and days of suffering. We report both because they donāt always agree. Eggs, for example, rank high on days of suffering per serving but lower on lives because one hen lays many eggs over her life. Each metric is available per serving, per kilogram, and at the national level, which supports three different decisions:
- Per serving (individual): what one person changes by cutting a specific product.
- Per kilogram (institutional): the impact behind procurement and wholesale purchasing.
- National level: which products, if replaced at scale, would spare the most animals overall.
The Headline Pattern
Across both countries, chicken, fish, and egg products cause the most suffering and take the most lives per serving. Dairy products consistently ranks lowest. For a concrete anchor: chicken luncheon meat sits at more than 9 days of suffering per serving, the highest of the 97 products analyzed. This pattern reflects short lifespans and small body sizes relative to the number of animals involved.
The Climate Trap Worth Flagging
Many institutions now optimize food purchasing for carbon. That framing makes switching from beef to chicken or fish look like an obvious win. But chicken and fish come from muchĀ smaller animals, so many more individuals are affected per kilogram. An intervention that reduces emissions can increase animal suffering. APIS is built to let advocates put both numbers in front of decision-makers at once ā climate data to get in the room, animal impact data to guide what actually gets purchased.
Cross-Country Validation
The U.S. and U.K. estimates draw on entirely separate national data sources ā the U.K. figures use the 2019ā23 National Diet and Nutrition Survey and DEFRA supply data ā yet they land in broadly the same place across every product category. Two independent models agreeing is one of the more reassuring results here. The main divergence is in fish, which is expected: the two countries eat different species, and species vary widely in body size, lifespan, yield, and whether theyāre farmed or wild-caught.
What Changed In The Methodology (And Why Estimates Dropped)
The 2026 figures are generally somewhat lower than previous estimates. This isnāt because animals are suffering less ā itās because we corrected modeling errors and updated the data. In the spirit of transparency:
- Corrected days-of-suffering formula. The old formula multiplied indirect impacts (like feed fish and pre-slaughter mortality) by the primary animalās lifespan. A wild-caught anchovy used as feed suffers for one day regardless of what eats them. The corrected formula treats indirect impacts separately.
- New dairy allocation factor. Following the International Dairy Federationās standard, we now attribute 88% of a dairy cowās lifetime impact to milk and 12% to beef.
- Corrected dairy loss proportion. Byproducts like skim milk and buttermilk were previously treated as lost. Theyāre actually consumed, so the model now uses milk-solids ratios.
- Molting in the U.S. egg model. About 37% of U.S. egg farms extend hensā laying cycles across two seasons. Molted hens lay more eggs per day of life, which lowers the estimated suffering per kilogram of eggs.
- Updated data sources. U.S. consumption now uses the 2021ā23 NHANES survey, supply figures use the latest USDA Loss-Adjusted Food Availability data, and demand elasticities for beef, pork, and chicken use retail scanner data.
The two largest shifts were in pork (fewer feed-fish deaths attributed to it, after EUMOFA revised feed-fish allocation) and eggs (the molting correction).
How You Might Use It
- Alt-protein R&D and investment: identify which products offer the greatest potential to reduce suffering, as an animal-welfare lens to pair with market and feasibility considerations.
- Institutional and government advocacy: APIS estimates have already informed plant-based procurement policies in several major U.S. cities.
- Individual decisions: help reducetarians cut the highest-impact products first.
- Organizational strategy: ourĀ process-tracing case study shows how APIS data shaped Anima Internationalās pivot toward cage-free campaigns.
Explore It
The resource, full methodology, and raw data (downloadable from our OSF project page) are all linked here:Ā https://faunalytics.org/animal-product-impact-scales-2026-updated-u-s-estimates-and-the-u-k-s-first-animal-product-impact-data/
Hi Jienna. Thanks for the update.
Have Faunalytics and Bryant Research (@Chris Bryant šø or @Richie?) considered excluding effects on wild animals (such as wild fishes used for feed)? I am very uncertain about whether expanding agricultural land increases or decreases the welfare/suffering/happiness of soil invertebrates, and I believe effects on these may be much larger than the effects on farmed animals and wild animals used for feed.
I also do not know whether decreasing the catch of wild fishes for feed increases or decreases animal welfare. Michael St. Jules concluded "decreasingĀ the catch of wild aquatic animals for feed is likely to decreaseĀ aquaculture and increaseĀ insect farming, but has unclear effectsĀ on brine shrimp nauplii and other live feed".
Hi @Vasco Grilošø, thanks for your comment!
We do calculate both "total" (including feed fish and pre-slaughter mortality) and "direct" (excluding feed fish and pre-slaughter mortality) numbers of animals and days of suffering, and that data is in the sheets on the OSF page!
We also differentiate fish by species which are commonly farmed (e.g. salmon) vs. species which are commonly wild-caught (e.g. tuna), and wild-caught species get 1 day of suffering per animal life (i.e. the day they get fished).Ā
As feed, wild-caught fish are overwhelmingly used in aquaculture (92%), and then a little bit in pigs (5% IIRC) and poultry (3% IIRC), so those are the species driving that number (cows are not commonly fed fish.)
On the broader point about the counterfactual impact on wild animal suffering -- I'm afraid that is mostly a factor that this model does not consider, but is a great candidate for a feature to add in future versions.Ā
My ill-informed read is that the ultimate net impact on wild animal suffering would be extremely difficult to predict. For example, tuna (mostly wild-caught) are carnivorous, so the counterfactual there is that they'd be eating many times more fish -- but then some of those fish also eat other fish! So the whole thing feels quite intractable quite quickly, but as I say, I don't know a ton about this topic, so maybe that's a cop-out...!
I hope this was helpful, and I'd be happy to be corrected on the tractability of considering net impact on wild animal suffering!
Hi Chris. Thanks for the clarifications. I agree including more effects on wild animals would be difficult. Note my question was whether you have "considered excluding effects on wild animals (such as wild fishes used for feed)". You might have read "including" instead of "excluding". Would it be better to exclude effects on all wild animals? It could be that the wild fishes used for feed would have suffered more if they were not caught.