Building Interventions in Low- and Middle-Income Countries

Summary: This post discusses the limitations of Western models of animal advocacy in lower- and middle-income countries (LMICs). It talks about how FWI tackles these challenges and concludes with some advice for others interested in this field.

About six years ago, my co-founder, Haven, put on a much nicer shirt than usual (we had to buy it the weekend before), stood in front of a modest audience, and announced Fish Welfare Initiative. He outlined a clear vision:

“Now you all have heard of the oft-mentioned Against Malaria Foundation. It's a charity hyper-focused on one extremely important issue: providing anti-malaria bed nets to people who need them most. It's arguably the most effective charity out there in the global poverty space. Imagine if we could run the same sort of model for animals.”

Haven introducing Fish Welfare Initiative in 2019.

Initially, our plan seemed beautifully simple: six months of desk research, a six-month pilot, then scaling rapidly—probably through corporate engagement or direct distribution of welfare technology like aerators. This was broadly the suggestion from AIM and was endorsed by the wider animal movement at the time.

Then we actually tried it.

The pilot didn’t work. NGO partners couldn’t connect us with farmers, farmers who did participate didn’t trust us, and the scientific data we needed barely existed. We had assumed farms followed standardized practices, that corporate supply chains could easily influence conditions, and that basic resources like water were reliably accessible year-round. After about 5 months of sweat, we decided to pull the plug on the pilot.

This was our first real lesson in something we'd soon become very familiar with: Western-centric models wouldn’t solve fish welfare in India.

Framing the Issue

If you're reading this, you’re likely aware of the massive scale of animal farming in low- and middle-income countries (LMICs)—about 84% of global fisheries and aquaculture production comes from these regions. You might also know about the common challenges: a smaller animal advocacy movement, informal markets, and weak policy enforcement.

At FWI, we think the root issue is deeper—a fundamental lack of infrastructure:

• Corporations lack the infrastructure to control their supply chains (e.g., not necessarily knowing where their products are coming from)

• Farmers lack the infrastructure to control their production systems (e.g., a lack of training and unpredictable profits that make it hard to take financial bets on things like welfare technology)

• Governments lack the infrastructure to regulate production and markets (e.g., 90% of agricultural jobs are part of the informal economy, meaning they are operating outside of formal government systems)

For us, engaging with corporations has mostly been a dead end, despite multiple attempts. We've learned that any successful intervention must closely align with local realities. For instance, our Alliance for Responsible Aquaculture is partly a response to the inadequate water-quality monitoring infrastructure available to farmers in Andhra Pradesh. Without such infrastructure, most farmers remain unaware of their fishes' actual conditions until issues become visibly severe (for example, farmers are only addressing low dissolved oxygen when fishes are already desperately gasping for air). 

Our Current Approach

We drew significant inspiration from our peers in the agricultural development space. Their use of farmer-producer organizations, accessible technologies, and strategic collaboration with government agencies served as compelling case studies for how to succeed in Indian aquaculture.

We also started to read more into the global development space’s concepts for making change. For example, the stage-based learning system, developed by Innovations for Poverty Action (IPA), outlines discrete stages all interventions go through (Ideate, Refine, Prove, Adapt, and Scale) and each stage requires different types of learning and evaluation. In early stages, the goal isn’t to prove impact in a rigorous study (as we were trying with our pilot). It’s to understand the problem, test assumptions, and build something that actually fits the context. Only once those basics are in place should we worry about scaling.

We now think that building real change in LMICs means two things:

1. Action: Getting in the field, testing things, working with farmers. This is hard, messy, and essential.

2. Research: Learning what’s actually needed and what’s actually possible, then iterating based on that. 

FWI operates through three main departments, each deeply integrating action and research:

• Programs Department (ARA Program): Our main department for “action”, working directly with farmers to develop our understanding and have short-term impact.

• Research and Development Department: Our main department for “research”. This team conducts in-depth, rigorous studies such as our Dissolved Oxygen tolerance research, fortified feed trials, and the development of holistic welfare assessment protocols. These activities directly inform practical interventions and expand our evidence base.

• Exploratory Programs Department: Our “in-the-middle” department. They do quick and dirty in-field work into big, uncertain ideas like promoting stunners in India or setting up a WhatsApp chatbot to give farmers advice.

The Alliance for Responsible Aquaculture (Praveen) working with consultants for the Exploratory Programs Department. Having both action and research actors on the team allows for shared knowledge and resources.

As such, we have built our work around the belief that meaningful change in LMICs requires direct, on-the-ground work combined with robust, actionable research.

Our Suggested Approach

Here are the four suggestions we have for setting up animal welfare interventions in LMICs, based on our experience:

1. Identify what stage you are in: Adopt frameworks like IPA’s stage-based learning to structure your approach appropriately.

2. Learn from agricultural development: Draw on proven models like farmer-producer organizations and accessible technologies.

3. Engage directly with the field early: Begin practical testing and collaboration with local stakeholders from the start.

4. Embrace iterative research and improvement: Commit to continuous learning and adapt strategies based on direct feedback and evidence.

Taking it a Step Further

We are pretty serious about this model of making change. So serious, in fact, that I (Tom) am looking into branching into this entirely. You may have seen our recent blog post announcing that I am going to be leaving FWI as a formal paid team member. Over the coming months, I will be looking to lead a research project exploring successful LMIC strategies from agri-development that could apply to farmed animal welfare.

If you're curious about this future direction, check out this EA Forum post. For those especially interested, we also have an expression of interest form to get involved.

At FWI, we have moved from trying to scale models prematurely to building the foundations that a model needs to work. We don’t have all the answers, but we’re asking better questions, and we’re optimistic that we’re helping to lay the groundwork for a new kind of animal advocacy, built to work where most farmed animals live.