Most published animal welfare research is carried out by universities in labs.
I kind of assumed that this was a necessity in some way. Either stuff can’t get published if you’re not affiliated with a university, or universities just have the necessary rigour to get things done. I just assumed there’s a whole host of reasons why this had to be the case, but I think those are mistaken.
And as a result, I think that when you start to look at universities, it starts to look strangely expensive to get things done. Things have a much longer time scale than is useful. And the facilities that they have to set up are often not good indicators of what reality looks like outside of the lab.
I think that high-quality welfare work can be done outside of universities and in fact, I think a huge amount of welfare work is done outside universities on farms but just isn’t actually published and available in the literature. It just needs to be brought to the surface.
I want to see if it’s possible for the animal welfare movement to rethink the model of research work and actively turn farms into welfare labs. I don’t think this is necessarily possible across the board, but I do think there are quite a few engaged farmers who would be willing to collaborate on this. They already have a really great setup and may already be doing a lot of this work anyway.
How do we actually get data from farms?
There are probably a few things we could do here. Like through insurance or banks or auditors, providing this data in an anonymised way to various authorities could become a requirement of receiving approval for loans, coverage, etc. We could work through unions like the National Farmers Union, where there might be a collective good for them to share the data and share that knowledge. We could pay them directly for the data and subsidise the costs to facilitate data sharing. We could set up our own animal welfare certification body that requires data sharing.
We could start with data in places where there’s already lots of data already—aquaculture and salmon in particular seem to be doing really well. There are companies like AgriGates, which is like data infrastructure for Precision Livestock Farming.
There are already places that are tracking lots of data, like slaughterhouses that already track a bunch of data from a bunch of different farms. They track things like body condition scores for the animals that they receive and they do this to pay farmers depending on the scores. So we could link this to on-farm condition data sets. Although we might be skeptical about the actual quality of this data, we can maybe get it from freedom of information requests or public records.
Can we turn them into Welfare Labs?
What is it that we’d actually want to do? I think this would start by identifying pertinent welfare indicators and exploring what a sort of coalition with farms would look like. It’s also worth trying to tap into some of the funding sources that universities are already accessing.
We probably want to do lots of preference testing or motivation testing or enrichment testing. For example, variable lighting: Welfare researchers have already been paid by Tyson to do variable lighting tests for broilers. This kind of work is already happening, it’s just not getting published in the traditional academic sense.
It’s important to talk with farmers, and in particular, farmers whom we think could be really engaged, like Vertical Oceans. Farmers in China could also be amenable to this—they are interested in precision farming, and they have subsidies for installing PLF tech.
The RSPCA wants to monitor their certification scheme using precision welfare tech. Can we just find out what it is they actually want and then just build it ourselves? There’s no reason why this needs to go through a university.
Why this makes sense
When you think about it, farmers are already running experiments all the time, including welfare inputs and/​or outcomes. They’re testing different lighting regimes, different feeding schedules, different housing configurations. They’re collecting data on mortality, feed conversion, growth rates. The only difference is that they’re not writing it up for peer review.
And actually, their data might be more valuable than university data in many ways. It’s collected under real commercial conditions, with commercial genetics, at commercial scale. A university study with 500 birds doesn’t necessarily tell you what will happen with 50,000 birds in a real barn.
The timeline issue is also crucial. By the time a university study is designed, funded, approved by ethics committees, run, analysed, written up, and published, we’re generally looking at 3-5 years minimum. We need answers now. If we can create a system where farmers can share what they’re learning in real time, we could accelerate welfare improvements dramatically.
I think this is one of those ideas that seems radical at first but then becomes obvious when you think about it. Of course we should be learning from what’s actually happening on farms. Of course farmers should be partners in welfare research, not just subjects of it. The infrastructure is already there, the data is already being collected, we just need to connect the dots.
Turning Farms into Welfare Labs
Most published animal welfare research is carried out by universities in labs.
I kind of assumed that this was a necessity in some way. Either stuff can’t get published if you’re not affiliated with a university, or universities just have the necessary rigour to get things done. I just assumed there’s a whole host of reasons why this had to be the case, but I think those are mistaken.
And as a result, I think that when you start to look at universities, it starts to look strangely expensive to get things done. Things have a much longer time scale than is useful. And the facilities that they have to set up are often not good indicators of what reality looks like outside of the lab.
I think that high-quality welfare work can be done outside of universities and in fact, I think a huge amount of welfare work is done outside universities on farms but just isn’t actually published and available in the literature. It just needs to be brought to the surface.
I want to see if it’s possible for the animal welfare movement to rethink the model of research work and actively turn farms into welfare labs. I don’t think this is necessarily possible across the board, but I do think there are quite a few engaged farmers who would be willing to collaborate on this. They already have a really great setup and may already be doing a lot of this work anyway.
How do we actually get data from farms?
There are probably a few things we could do here. Like through insurance or banks or auditors, providing this data in an anonymised way to various authorities could become a requirement of receiving approval for loans, coverage, etc. We could work through unions like the National Farmers Union, where there might be a collective good for them to share the data and share that knowledge. We could pay them directly for the data and subsidise the costs to facilitate data sharing. We could set up our own animal welfare certification body that requires data sharing.
We could start with data in places where there’s already lots of data already—aquaculture and salmon in particular seem to be doing really well. There are companies like AgriGates, which is like data infrastructure for Precision Livestock Farming.
There are already places that are tracking lots of data, like slaughterhouses that already track a bunch of data from a bunch of different farms. They track things like body condition scores for the animals that they receive and they do this to pay farmers depending on the scores. So we could link this to on-farm condition data sets. Although we might be skeptical about the actual quality of this data, we can maybe get it from freedom of information requests or public records.
Can we turn them into Welfare Labs?
What is it that we’d actually want to do? I think this would start by identifying pertinent welfare indicators and exploring what a sort of coalition with farms would look like. It’s also worth trying to tap into some of the funding sources that universities are already accessing.
We probably want to do lots of preference testing or motivation testing or enrichment testing. For example, variable lighting: Welfare researchers have already been paid by Tyson to do variable lighting tests for broilers. This kind of work is already happening, it’s just not getting published in the traditional academic sense.
It’s important to talk with farmers, and in particular, farmers whom we think could be really engaged, like Vertical Oceans. Farmers in China could also be amenable to this—they are interested in precision farming, and they have subsidies for installing PLF tech.
The RSPCA wants to monitor their certification scheme using precision welfare tech. Can we just find out what it is they actually want and then just build it ourselves? There’s no reason why this needs to go through a university.
Why this makes sense
When you think about it, farmers are already running experiments all the time, including welfare inputs and/​or outcomes. They’re testing different lighting regimes, different feeding schedules, different housing configurations. They’re collecting data on mortality, feed conversion, growth rates. The only difference is that they’re not writing it up for peer review.
And actually, their data might be more valuable than university data in many ways. It’s collected under real commercial conditions, with commercial genetics, at commercial scale. A university study with 500 birds doesn’t necessarily tell you what will happen with 50,000 birds in a real barn.
The timeline issue is also crucial. By the time a university study is designed, funded, approved by ethics committees, run, analysed, written up, and published, we’re generally looking at 3-5 years minimum. We need answers now. If we can create a system where farmers can share what they’re learning in real time, we could accelerate welfare improvements dramatically.
I think this is one of those ideas that seems radical at first but then becomes obvious when you think about it. Of course we should be learning from what’s actually happening on farms. Of course farmers should be partners in welfare research, not just subjects of it. The infrastructure is already there, the data is already being collected, we just need to connect the dots.