This is a new episode of the GiveWell Conversations podcast. You can listen to the episode or read a summary of the conversation below.
GiveWell has built its reputation on rigorous research—analyzing randomized controlled trials, building cost-effectiveness models, and reviewing monitoring data to identify cost‑effective ways to save and improve lives.
In an effort to supplement this desk research and make better decisions, we’ve been working to gather more information directly from the people who live and work in the countries where we fund programs.
In this episode, GiveWell CEO and co-founder Elie Hassenfeld speaks with Principal Researcher Alex Cohen about GiveWell’s work to gather local insights to check our assumptions and figure out what we might be missing.
Listen to Episode 21: Testing Our Assumptions through Local Insights
Elie and Alex discuss:
- Testing key hypotheses about the data GiveWell relies on: We’re working to improve the inputs in our decision making. This includes checks on coverage data, more information about how programs work in practice, and assessing whether estimated program effects are plausible. These efforts have already identified discrepancies between coverage surveys and other areas for improvement.
- Addressing the limitations of global health data: GiveWell depends on credible data for its research, yet global health and development data is quite limited. For example, basic measurements like child mortality rates rely on household surveys conducted only every five years. We’re employing multiple approaches to address these limits: funding independent survey firms; conducting site visits; hiring local consultants; and strengthening networks with government officials, implementing organizations, and other funders.
- Balancing the trade-offs between local work and desk research: Desk research will continue to make up the vast majority of our work. We believe that complementing that research with additional information we gather from local sources could meaningfully improve our grantmaking. We expect to dedicate around 5% of our research team’s time and around 1% of our total grantmaking to these efforts, which we believe will have an outsized impact.
By prioritizing efforts to learn from people in the places where we fund programs, we hope to better understand how programs are being implemented, identify bottlenecks, and more. We believe that incorporating this information will improve our decision-making and our work to help people as much as we can. It provides checks on our primary models, increases confidence in our conclusions, and could highlight where we might be missing something important.
Visit our All Grants Fund page to learn more about how you can support this work, and listen or subscribe to our podcast for our latest updates.
This episode was recorded on December 23, 2025 and represents our best understanding at that time.
