It is always a bit magical when a conference turns into a conversation—and then into a movement. That’s what it felt like at this year’s Economic Science Association (ESA) Africa Conference hosted by Busara in Nairobi. The conference set something important in motion—a step toward ensuring that research in the Global South is shaped by the people it serves.
We came from different organizations—3ie and Busara, but found ourselves telling a shared story: one of evidence grounded in context, tools shaped by trust, and voices from the Africa and MENA region desiring to lead.
In a room filled with researchers, implementers, and government partners, Thomas Kelly, Geetika Pandya, and Abdelrahman from 3ie initiated the pre-conference activities by hosting a half-day training on evidence synthesis and AI. They added a different dimension to our participants' perception of AI and encouraged everyone to think of it in a new light—a tool that can enable evidence to be faster, smarter, and if we dare to make it, more inclusive. What would it mean for research if we could systematically review and meta-analyze all the knowledge we have already procured? To enable it to serve us better and more importantly, faster?
3ie trained the participants on evidence synthesis using AI, and then showed everyone exactly what it would look like to make use of AI to support real-world policymaking, specifically in West Africa by presenting us with a case from WACIE (West Africa Capacity-building and Impact Evaluation), where 3ie used AI to provide rapid synthesis and evidence translation to help policymakers understand what evidence exists for specific policy questions.
The next day, participants got an opportunity to listen to panelists Rasha Attar, Benjamin Krause, and Geetika Pandya discuss the capability of AI and what we should be asking of it. They held a discussion on systems that learn and knowledge infrastructures, while emphasizing the essential role of the communities whose lived experiences must guide these systems. In the context of the Global South, this calls for a shift where research is not externally imposed, but rather led by scholars, institutions, and communities from the Global South itself. A significant learning from this panel discussion was that how useful AI is, depends on the questions we ask it and who we're trying to help with the answers.
Innovation is easy, impact is hard
We then had research presentations on the structures that shape the Global South.
Abdelrahman shared findings from a large-scale randomized controlled trial in Upper Egypt on parental pressure and youth labor market outcomes. The study was grounded in the experiences of over 20,000 job seekers, each navigating choices shaped by economics, family dynamics, gender norms, and aspirations for a better future.
The data told a compelling story, fathers often urged their children forward, while mothers tended to be more protective. Young men responded by widening their job search, taking more risks. In contrast, young women were more hesitant—often holding back, shaped by both social expectations and family concerns.
But the research revealed something deeper. When parents were engaged meaningfully and brought into the conversation, the outcomes began to shift. For instance, involving an educated mother in the process boosts a daughter’s likelihood of seeking work. It became clear that labor market decisions saw more influence at dinner tables than at job centers or classrooms. Sometimes, it is everyday conversations that shape outcomes.
Geetika also presented an examination of agricultural insurance across Kenya, Senegal, and India. Drawing on a synthesis of five impact evaluations, she highlighted a familiar challenge: while innovation in this area is achievable, delivering meaningful impact proves far more difficult. The interventions introduced a range of innovations, including picture-based apps for reporting crop damage, the use of satellite and ground-reference imagery to develop plot-level credit scores, and financial tools such as bundled credit and insurance. Although these efforts increased the awareness of insurance among smallholder farmers, they delivered limited measurable improvements in wellbeing outcomes such as food security and health. To enhance impact, policymakers must prioritize thoughtful intervention design that fosters both understanding and trust in insurance products and services.
Long-term partnerships
What mattered most at ESA Africa 2025 were the African voices leading the research and the focus on regional priorities, signaling a collective shift toward ownership of the development agenda. For years, conversations regarding development have been dictated from outside the region, but this conference marked a clear moment where the dialogue came from within.
This shift is especially critical now, amid budget cuts and the growing challenges faced by academia. Traditional research and policy models are showing their limitations, and within those gaps, collaboration is emerging as a sustainable path forward.
Recognizing this, 3ie and Busara have begun a long-term partnership, not just between themselves, but as part of a broader network of organizations that share a commitment to rigorous research and view evidence as a public good. Our goal is to build a space where the Africa and MENA region lead the research agenda, and where policy is proactive, grounded in evidence, informed by behavior, and shaped by the communities it serves.
So, what’s next?
More shared questions. More honest answers. And more spaces like ESA Africa, where evidence isn’t discussed in the abstract but is rooted in the lives and voices of those it’s meant to serve.
Because at the end of the day, evidence is not just about what works. It’s about who it works for—and who gets to decide.
