As for your post about EA & a lot of EA participants being in far away developed countries, I agree & my exposure to this line of thinking largely comes from the book The Anti-Politics Machine (https://en.wikipedia.org/wiki/The_Anti-Politics_Machine) describing previous failures in international development from a lack of local knowledge. It’s actually something I find interesting about EA interventions, as instead of trying to develop detailed local knowledge they work around the issue with things that require less local knowledge to work. It’s easier to support a one-time technical intervention like vaccines with like New Incentives that do not require continued follow-up throughout a person’s life or to just send cash with GiveDirectly. And I can’t say this is a bad or even sub-optimal approach as things like vaccines have such great benefits for such low cost & low risk of intervention failure. It more reveals how under funded everything is that things like vaccines & cash transfers to expectant mothers aren’t fully saturated leaving only interventions that could use more local context.
This is one of the most useful framings I’ve encountered for a tension I’ve felt but struggled to articulate clearly. James Ferguson’s argument that development interventions systematically depoliticize the problems they claim to solve maps directly onto what I observed in the field. The machine produces outputs, but the underlying power structures remain untouched. Your point about EA working around the local knowledge problem rather than through it is sharp, and I think largely correct. Vaccines and cash transfers are genuinely brilliant precisely because they minimize the surface area for implementation failure. But I’d push back slightly on the implication that this is sufficient as a long-term strategy. The interventions that are currently “not saturated” as you put it are not randomly distributed. They tend to cluster in exactly the domains where local context, political economy, and institutional relationships matter most: primary healthcare systems, land governance, nutrition policy, climate adaptation. At some point, EA will have to engage with the messiness it has been successfully avoiding. That’s not a critique it’s what I think the next frontier looks like. And it’s part of why I think people with field experience in these contexts have something genuinely useful to contribute to the conversation, rather than just being recipients of well-designed interventions.
As for your post about EA & a lot of EA participants being in far away developed countries, I agree & my exposure to this line of thinking largely comes from the book The Anti-Politics Machine (https://en.wikipedia.org/wiki/The_Anti-Politics_Machine) describing previous failures in international development from a lack of local knowledge. It’s actually something I find interesting about EA interventions, as instead of trying to develop detailed local knowledge they work around the issue with things that require less local knowledge to work. It’s easier to support a one-time technical intervention like vaccines with like New Incentives that do not require continued follow-up throughout a person’s life or to just send cash with GiveDirectly. And I can’t say this is a bad or even sub-optimal approach as things like vaccines have such great benefits for such low cost & low risk of intervention failure. It more reveals how under funded everything is that things like vaccines & cash transfers to expectant mothers aren’t fully saturated leaving only interventions that could use more local context.
This is one of the most useful framings I’ve encountered for a tension I’ve felt but struggled to articulate clearly. James Ferguson’s argument that development interventions systematically depoliticize the problems they claim to solve maps directly onto what I observed in the field. The machine produces outputs, but the underlying power structures remain untouched.
Your point about EA working around the local knowledge problem rather than through it is sharp, and I think largely correct. Vaccines and cash transfers are genuinely brilliant precisely because they minimize the surface area for implementation failure. But I’d push back slightly on the implication that this is sufficient as a long-term strategy. The interventions that are currently “not saturated” as you put it are not randomly distributed. They tend to cluster in exactly the domains where local context, political economy, and institutional relationships matter most: primary healthcare systems, land governance, nutrition policy, climate adaptation. At some point, EA will have to engage with the messiness it has been successfully avoiding.
That’s not a critique it’s what I think the next frontier looks like. And it’s part of why I think people with field experience in these contexts have something genuinely useful to contribute to the conversation, rather than just being recipients of well-designed interventions.