Turning Farms into Welfare Labs

Aaron, this is a great idea. I strongly agree that bringing research closer to commercial farms is essential if we want findings that actually reflect what happens in practice, with much of what is produced in research settings (even those that try to mimick commercial practice) suffering from what we call the 'healthy farm effect'. Commercial data capture the full messiness of real production, which is why it is so valuable.

External validity is not the only thing missing in welfare science. Most of what we know about animal welfare at commercial scale comes from single visits to farms, essentially a photograph of what happens. We need the video (longitudinal research). We need to know when different welfare problems start, how long they last, and how many animals are affected. In short, we need an epidemiology of animal welfare, and commercial farms are the only place where that can happen.

That said, for this to work, I believe a few things are needed:

• Independence between funding and publication. In my experience working with industry-funded groups, there are often contractual or simply informal pressures (e.g. anticipation of future fundign) that discourage publishing unfavorable results . Agreements need to explicitly guarantee the right to publish regardless of outcome.
• Independent analysis. Data collection can happen on farms, but analysis should be conducted or audited independently to reduce potential biases.
• A clever design of incentives. Farmers should benefit from participating, but incentives shouldn't promote selective reporting.
• Standardization. The power of research on commercial farms will come from collecting consistents data and indicators across many farms, enabling large scale analyses with the proper statistical power and validity. This needs pre-defined protocols, clear definitions of variables and methods, preregistration where possible, and transparent data access. Without this, farm data risk being too messy or selectively reported.

With these safeguards, I believe research on farms will be of immense value. Farms are already generating huge amounts of data as you mentioned, one challenge now is creating the proper systems to use it .

Aaron!

Thanks for posting :) I’m coming at this as someone who spent a lot of time running on-farm research at Fish Welfare Initiative (and planning to do more through my new charity). I broadly agree, but I’d add a few caveats:

The core issue with farms as “welfare labs” is heterogeneity (variability). Especially in LMICs, farms are messy, uncontrolled environments where confounding variables easily creep in. That creates a lot of statistical noise. If you’re aiming for high certainty, farms can make that difficult.

Relatedly, on-farm research makes it harder to isolate specific effects. You mention the benefit of insights collected “under real commercial conditions”, and I agree that ultimately effectiveness in the real world is what matters. But there’s also strong value in isolating variables to understand mechanisms. If we’re testing whether pigs prefer straw or wood shavings, we may not want to simultaneously capture differences in how farmers manage those materials. Otherwise, when results don’t support our hypothesis, we don’t know whether we’re observing animal preference or management differences. That’s why the difference between efficacy (does it work in controlled conditions?) and effectiveness (does it work in real-world conditions?) is useful.

So in sum: on-farm research is a valuable tool, but it can’t replace controlled research, and I’d hesitate to frame it as more valuable.

I also think two ideas may be getting conflated: monitoring existing farm conditions vs running experiments on farms. Both can be useful, but both need a clear use-case for the data. I agree, though, that there’s strong potential on both fronts.

Great post, Aaron — I completely agree with your framing of why the lab-to-farm leap feels overdue. Most published welfare research is tied up in universities and controlled settings, which often are not only expensive and slow but often miss how animals actually experience their environments on commercial operations. 

I love your emphasis on starting with engaged farmers — that is a low-friction entry point, especially because so much welfare-relevant data is already being collected in everyday farm management but never shared or analyzed. If even a handful of farmers were willing to share anonymized data, we could begin to extrapolate welfare trends across operations and turn those insights into sector-wide benchmarks that signal progress in a measurable way.

Crucially, this model can offer upside for participating companies and farms: a feedback loop with welfare experts that helps them refine practices, capture wins, and communicate them externally — for example, “Our data-driven changes reduced mortality (or time in intense pain) by 15% — here’s how.” That kind of concrete progress, grounded in real farm data, can add a powerful real-world layer to academic research and help translate existing knowledge into practice at scale.

Some half-baked thoughts that you probably have already considered:

Definitely agree that welfare science could be moving much faster than it currently is, and part of that is tied to the research process in universities. But many of the restrictions that make the process slower do have benefits wrt to rigour. 

In an ideal world, the research produced by universities is trusted because it passes through a structured process designed to limit bias and enhance rigour. A good study is typically specified in advance: what intervention will occur, how outcomes will be measured, and how they will be analysed. Measurement uses agreed definitions so results are comparable across sites. The protocol is documented and deviations are recorded. Data are retained and can be inspected. Researchers are accountable to supervisors, ethics committees, and institutional reputation, which creates penalties for selective reporting. 

After that, peer review functions as an additional filter. Independent reviewers, who (hopefully) have no stake in the result, check whether alternative explanations could account for the findings, whether statistics are appropriate, and whether the conclusions match the data. Weak causal claims can be weeded out at this stage, and appropriate qualifications made. 

On farms, how much of this holds? I imagine that the same person might feasibly implement the change, observe the outcome, and possibly even benefit from their interpretation. Management conditions might change simultaneously, unsuccessful trials might be rarely recorded in a standardised way, and there may be no routine adversarial check of the inference by a domain expert. That makes it difficult for third parties to judge whether an alleged welfare improvement genuinely reflects the supposed intervention.

Collecting farm data is feasible, but achieving comparable credibility requires some mechanism(s) for maintaining rigour, including but not limited to: predefined methods, independent verification, and systematic challenge of the conclusions by experts. 

My two genuine (not-leading) questions would be: 

1) How feasible is this in the farm environment? This means not only implementing the mechanisms to ensure rigour, but also ensuring that key stakeholders trust the process. Even if your research is sound, if key stakeholders won't trust your results because you did it in a farm and not a lab, then you definitely have to take that into consideration rather than scoff at it.

2) Once you implement those mechanisms for rigour, are you still significantly quicker and more streamlined than university research? (If the answer to this is yes, one would wonder why universities aren't able to do the same. But obviously there are non-research related factors influencing university research).

I imagine that there's a trade-off between rigour and resources, where you the more rigour you want, the more resources (including time) you need. So a third bonus question is, how much rigour is enough?

Cool idea. I assume you're tracking Innovate Animal Ag already; they might be interested in this

Academic studies are definitely slow, but 3-5 years strikes me as extremely slow, even for academia. 

I'm generally on board with what you're describing but I wonder whether there's also opportunities to work better with academia? Like if you're the one providing the funding, you might be able to negotiate with them and keep them accountable to timelines. There's probably also lots of variability between academics so you can identify the ones that are capable of executing well and quickly, and then work with them. 

Thanks for the interesting post, Aaron. It reminded me of the Pain Atlas Project from the Welfare Footprint Institute (WFI). I shared your post with Wladimir Alonso, and Cynthia Schuck-Paim from WFI.

Executive summary: The author argues that animal welfare research does not need to be primarily university-and-lab-based, and that the movement should “turn farms into welfare labs” by surfacing and sharing high-quality welfare data already generated under commercial conditions.

Key points:

1. The author thinks universities look “strangely expensive,” slow (often “3–5 years minimum”), and sometimes unrepresentative of real farm conditions, and that these features are not necessary for rigor.
2. The author believes “a huge amount” of welfare-relevant work already happens on farms but is not published or accessible in the literature.
3. The author suggests multiple routes to obtain farm data, including tying anonymized data sharing to insurance, bank loans, audits, unions (e.g., the National Farmers Union), direct payment/subsidies, or a certification body that requires data sharing.
4. The author proposes starting with sectors that already collect lots of data (they mention aquaculture/salmon and “Precision Livestock Farming” infrastructure, including AgriGates).
5. The author notes slaughterhouses already track cross-farm metrics (e.g., body condition scores used for payment) and suggests linking these to on-farm datasets, potentially via FOI/public records despite concerns about data quality.
6. The author envisions farm-based welfare research focusing on welfare indicators and applied tests (preference, motivation, enrichment; e.g., variable lighting trials for broilers allegedly funded by Tyson) and argues this work could be built outside universities, including by aligning with farmers and certification schemes (e.g., RSPCA monitoring via precision welfare tech).

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.