Listening to LARTES and ARED share lessons from more than 30 years of practice at the recent Daara retreat in Senegal, I was reminded that working with government is not a box to tick at the end of a project. It is a long-term commitment that ultimately determines whether education innovations translate into system-level impact. What stayed with me most was their conviction that impact happens when governments are engaged early, treated as genuine partners in learning, and supported with evidence that is practical and immediately usable. This blog is an expanded reflection on the insights ARED and LARTES shared at the retreat – and the lessons I am continuing to sit with.
Engage government early and continuously
One of my strongest takeaways was the importance of beginning with government, not ending with it.
ARED and LARTES showed that engagement starts at problem definition – before research designs are finalized, before implementation begins. Ministries are not simply informed; they help shape the questions and influence the methods. That early alignment with national priorities builds legitimacy and reduces friction later.
I was particularly struck by how engagement extends beyond senior leadership. Technical departments, regional actors, and academic institutions are involved as co-designers. This approach shifts the dynamic entirely. When governments help define the problem and co-create the evidence, they are far more likely to trust the findings and use them. Ownership grows naturally.
Trust is built through complementary roles and credibility
Another lesson that stayed with me is the clarity of roles within the ARED-LARTES-Ministry partnership.
ARED brings deep field presence and long-standing credibility with teachers and education authorities – credibility reinforced by national and international recognition, including the 2025 Yidan Prize for Education Development awarded to Mamadou Amadou Ly. LARTES contributes methodological rigor, strong data systems, and analytical expertise. The Ministry provides policy direction, institutional authority, and pathways to scale. Beyond this, the government brings critical technical leadership, contextual knowledge, and strategic direction that shape how reforms are designed, sequenced, and embedded within national systems. Their expertise ensures that innovations are not only authorized, but technically sound, politically feasible, and aligned with long-term sector priorities.
No one tries to replace the other. There are no parallel systems. Instead, there are joint governance structures, shared tools, and validated terms of reference. Watching this dynamic reinforced an important insight for me: trust is not built through rhetoric. It is built through clarity, respect for mandates, and delivering on your role consistently.
Producing knowledge that is immediately useful
I also reflected on what they described as producing “useful knowledge.” This idea resonated deeply.
Research is not conducted for reports that sit on shelves. It is designed to influence classroom practice, teacher training, remediation strategies, and system monitoring in real time. Evidence is generated alongside ministry and regional actors, and findings are used to refine pedagogy and materials as implementation unfolds.
What I learned here is that uptake accelerates when evidence improves daily practice. When teachers see data helping them teach better, and when ministries see it strengthening decision-making, research stops feeling like external accountability and becomes a shared tool for improvement.
Scaling as institutional anchoring
One of the most powerful shifts in my thinking was around scaling. ARED’s journey made it clear that scaling is not replication. It is institutional anchoring.
Innovations gain traction when they align with existing reform agendas – such as bilingual education or structured remediation frameworks under national programs like MOHEBS. Governments are more receptive when interventions arrive fully formed: quality-assured materials, trained personnel, monitoring systems, and sustained technical accompaniment. Trust is built gradually through consistent delivery, adaptability, and responsiveness to system constraints such as staffing, financing, and administrative processes.
What stood out to me is the humility embedded in this approach. Adoption happens when innovation fits within the system’s logic and capacity. Systems should not have to bend around innovation; innovation must adapt to the system.
Government as a learning partner
Perhaps the most meaningful lesson for me was the reframing of government as a learning partner.
Through action research, longitudinal data, and structured feedback loops, ministries are not passive recipients of evaluation findings. They are co-interpreters. Evidence becomes a shared resource for strengthening the system over time.
This approach builds internal capacity – not just to use data for one intervention, but to embed continuous improvement into the system itself. It reminded me that real partnership is about strengthening institutions long after a specific project ends.
What I am taking forward
As I left the retreat, I found myself reconsidering how I think about innovation and impact. The experiences of ARED and LARTES reinforced a simple but profound lesson: lasting change in foundational learning is achieved not by working around governments, but by working patiently, strategically, and respectfully with them.
The system is ultimately what sustains change.
And that requires time, trust, and a willingness to learn together.
Dr. Lydia Chege
Lead Ecosystem Building & Government Engagement
Zizi Afrique Foundation, Kenya
English 
French