What Three Government Consultations Taught Us About Evidence, Power, and EdTech Policy in Africa 

There is a particular kind of insight that only emerges when you sit across the table from the government. Not from a survey, not from a policy document, not even from an interview — but from two full days of structured conversation with the technical directors, curriculum specialists, ICT leads, and senior officials who are actually responsible for educating millions of children.

Over the past months, the Africa EdTech Research Initiative (AERI) — a collaboration between EdTech Hub, the Association for the Development of Education in Africa (ADEA), and Education Sub Saharan Africa (ESSA) — has been doing exactly that. We held country-level consultations in Naivasha (Kenya), Saly (Senegal) and Lilongwe (Malawi), bringing together multi-sectoral government teams to explore a deceptively simple question: how do Ministries of Education in Sub-Saharan Africa Coordinate EdTech evidence, aligned with national priorities for use in policy decisions, particularly those that affect foundational learning?

What we found was more nuanced — and more instructive — than we anticipated.

The evidence exists

The first thing to understand is that governments are not passive or evidence-light environments. Across all three countries, Ministries of Education, ICT directorates and related agencies collect substantial data, draw on research during program design, and make decisions informed by implementation experience. Evidence use is happening.

But it is largely invisible.

Governments are under no obligation to publish their internal evidence or to make their decision-making processes visible to external actors. There are studies that inform budget decisions, evaluations that shape program revisions, and internal analyses that guide pilot designs. Yet, these rarely surface in the research literature or in partner-facing communications.

For researchers and development partners accustomed to looking for published evidence, this creates a systematic blind spot. The absence of visible evidence is not the same as the absence of evidence.

This matters enormously for how we think about the problem. The challenge is not to introduce evidence into government decision-making. It is to understand how evidence already moves through government systems — and to build coordination infrastructure that makes that movement more visible, more aligned and more durable.

The AERI project is designed to investigate both the formal and informal coordination mechanisms across ministries of education in Kenya, Malawi, and Senegal. The assumptions for the AERI project are not that coordination mechanisms do not exist, but through the study, which has allocated finances to implement and disseminate tech- enabled Foundational Literacy and Numeracy (FLN) research, in the three countries, EdTech fellows will observe the coordination mechanisms used across the evidence ecosystem, from generation to use in decision making. 

Formal mechanisms exist, informal ones govern

A related finding emerged consistently across all three consultations: the gap between documented and practiced coordination.

Each country has formal coordination mechanisms on paper — inter-agency frameworks, evidence protocols, structures for cross-ministerial collaboration. These exist in national strategic plans and policy documents. What governs actual practice, however, is a parallel set of informal mechanisms: trusted individual relationships, ad hoc communication channels, and the influence of particular champions at particular moments.

This is not a failure of governance. It is a recognisable feature of how complex institutions function under resource and time pressure. But these informal mechanisms have profound implications for evidence use. When coordination depends on individuals rather than systems, it is fragile. When a champion leaves a ministry, a coordination mechanism can quietly collapse. When relationships sour between agencies, information stops flowing. The result — visible in all three countries — is what several participants described as islands of excellence: pockets of strong, evidence-informed implementation that exist because the right person is in the right role, but which do not scale or survive institutional transitions.

The implication for AERI’s research agenda is clear: understanding coordination means understanding both the formal architecture and the informal practice, and the distance between them. The study will look at how coordination actually happens, not how the process is delineated on paper.

The policy window problem is structural, not attitudinal

Perhaps the most important insight from the three consultations — and the one most frequently misunderstood in the development sector — is what we are calling the policy window problem.

National implementation plans in all three countries operate on three-to-five-year cycles, anchored in legal frameworks and budget allocations. These cycles create defined windows, during which external evidence can meaningfully influence policy: the strategy development phase, the review and revision process, and the budget formulation period. Outside these windows, evidence does not stop being created — it simply has nowhere to land. Incorporating it would require reopening processes that are legally and administratively closed.

This is not a sign that governments are resistant to evidence. It is the institutional logic of how government delivery works. A Ministry cannot simply redirect a programme budget mid-cycle because a new study has emerged, any more than a parliament can rewrite its annual budget in response to a journal article.

The consequences for the research and development community are significant. Evidence generated without reference to the policy calendar faces structural barriers to uptake that have nothing to do with its quality. Researchers who invest years in a rigorous evaluation and then present findings to a Ministry mid-cycle should not be surprised when the response is polite interest rather than policy change. The timing was wrong. The window had closed.

Compounding this is another finding: implementation and budgeting calendars are known internally to the government but are not shared externally. Partners may be involved in the early phases of strategy development, but once implementation begins, they are generally not in the room for budget allocation or operational planning. There is no outward-facing decision calendar in any of the three countries — no signal to the research community of when evidence would actually be useful and to whom.

The coordination problem runs in both directions

It would be easy to frame the evidence ecosystem challenge as a government problem — insufficient transparency, insufficient openness to external research. But the consultations made clear that the misalignment runs in both directions.

Because government EdTech priorities and research gaps are not readily visible externally, development partners and researchers frequently commission studies that governments do not need. Conversely, the evidence that governments are generating internally rarely reaches the research community in a form that could inform future work. The result is duplication — multiple actors generating overlapping evidence, none of it clearly connected to a shared understanding of what the sector needs to know.

Governments in all three countries named this directly. They spoke about the difficulty of tracking what partners are producing, the frustration of receiving findings that do not map onto current priorities, and the absence of any mechanism through which they could signal their evidence needs to the external research community.

This points to something the sector has struggled to build: a living, government-anchored research agenda for EdTech and foundational learning — one that reflects current national priorities, is visible to partners and researchers, and is updated as policy cycles evolve. Without it, the evidence ecosystem will continue to generate fragmented, poorly coordinated knowledge that struggles to reach the decision-making processes it is intended to inform.

Coordination is also an internal challenge

The coordination challenge does not stop at the boundary between government and partners. Within the government, it is equally acute.

The ministries and directorates that share responsibility for EdTech in education — Ministries of Education, Ministries of ICT, Science and Technology, teacher development agencies, curriculum bodies — often operate with siloed mandates. They share the same learners and the same teachers, but they do not always share evidence, coordinate on policy design, or align their implementation timelines. The same is true vertically: across early years, primary and secondary education, coherence in EdTech policy and evidence use is uneven.

In the absence of clear inter-agency coordination, fundamental questions go unanswered: whose mandate is it to collate and synthesise EdTech evidence across the system? Who is responsible for ensuring that a national policy on digital learning is consistently implemented at a sub-national level? Who carries institutional knowledge when a programme transitions from one political administration to the next?

In Kenya and Malawi,  the government’s honest assessments of their own EdTech portfolio was instructive. Several ongoing projects were implemented without a prior evidence base and without integrated research or monitoring to assess learning outcomes. This is not uncommon — implementation pressures, donor timelines and political visibility often drive decisions faster than evidence can be generated. But it does point to a systemic gap: the absence of an institutional framework for learning from implementation. Governments need structures that allow them to evaluate what they are doing, draw lessons, and carry those lessons forward across political transitions. Currently, those frameworks are largely missing.

The politics of evidence are real — and researchers must engage with them

One of the most candid exchanges across the three consultations concerned the politics of evidence use, and it is worth addressing directly.

Evidence does not land in a politically neutral environment. Governments are accountable to citizens and to political mandates. Evidence that frames their performance as inadequate — however accurate — creates genuine risk: it can be weaponised by political opponents, used to undermine public confidence, or deployed in ways that bear no relationship to the nuance of the original research. This is not a reason to suppress evidence. It is a reason to understand why governments sometimes respond to difficult findings with caution, deflection or silence.

Several practical implications follow. First, who presents evidence matters as much as what the evidence says. Credibility, trust and perceived intent are gatekeepers. Researchers perceived as adversarial, politically aligned with opposition actors, or pursuing institutional agendas are unlikely to see their findings taken up, regardless of methodological quality. Building sustained, trust-based relationships with both technical directors and senior leadership is not a soft skill — it is a prerequisite for evidence use.

Second, technical receptiveness is not the same as institutional uptake. Researchers frequently build strong working relationships with Ministry technical staff who are genuinely engaged with evidence. But it is senior management teams — operating with political considerations fully in view — who make final decisions. Evidence that impresses a technical directorate may never reach a decision-maker’s desk, or may arrive there translated in ways that change its meaning. Understanding the internal decision-making hierarchy, and engaging at multiple levels, is essential.

Third, and perhaps most importantly, how challenges are framed determines whether evidence is heard at all. Language that positions government as absent, deficient or failing — common shorthand in development sector communications — is not only often inaccurate, it is counterproductive. Governments are managing education systems at extraordinary scale, under conditions of fiscal constraint and systemic complexity that most research environments do not approximate. The appropriate posture for evidence-bringers is one of complementarity: research that builds on what the government is already doing, strengthens existing systems, and opens policy options rather than implying that current approaches are without value.

What a functional coordination mechanism would look like

Drawing across the three consultations, it is possible to sketch what a functional evidence coordination mechanism for EdTech and FLN would need to include.

• Government led and anchored. It would need to be government-anchored — not partner-led — so that it carries institutional authority and is trusted by the agencies whose behaviour it is trying to coordinate.
• National Research Agenda.  It would need to maintain a publicly accessible, regularly updated research agenda that reflects current government priorities and live research gaps, giving the external community a clear signal of what evidence would actually be useful.
• Policy Windows.  It would need to be explicitly timed to policy and budget cycles, creating defined entry points for external evidence rather than leaving this to chance. 
• Advocacy. And it would need an advocacy dimension — a capacity to make evidence legible and compelling to decision-makers who hold budget authority, not just to technical staff who are already persuaded.

In Kenya, some of the architecture for this already exists, if unevenly. The Ministry of Education’s Education Evidence Hub — embedded in the current National Education Strategic Plan (2023–2027) — was established with precisely this mandate: to receive, synthesise and disseminate education evidence, and to facilitate inter-agency learning. The hub exists, but it is not yet functioning at the scale or with the resources its mandate requires. But its existence is significant. It represents a government-owned coordination infrastructure that, properly supported, could anchor exactly the kind of ecosystem coordination the sector needs.

Several partner-led mechanisms — including Education Evidence 4 Action (EE4A), What Works Hub for Global Education (WWHGE) and the Education Development Partners Coordination Group — also play valuable coordination roles in Kenya. But as the consultations made clear, none of these currently operate at the system level: none has visibility into how evidence actually moves through internal government decision-making, or the authority to align that movement with national priorities. Filling that gap is what AERI is working to understand.

What comes next

The consultations in Lilongwe, Naivasha, and Saly were not endpoints. They were the beginning of the co-design process. In each country, governments set the agenda, mobilised multi-sectoral teams, and engaged substantively over two full days. The convenings succeeded in generating the trust, relational energy, and momentum needed for true collaboration. We co-created concrete action plans for the next 18 months:

• Defined the EdTech and Research Priorities: The governments named the challenges in EdTech and foundational learning in their own languages and with their own priorities. 
• Finalise Results Frameworks: The government tailored country-level results frameworks directly to the specific coordination blocks identified in each nation.
• Execute Ministry-Driven Research: We are deploying project resources to answer one high-priority EdTech/FLN research question defined entirely by each respective ministry.
• Observe the Ecosystem: As our teams work to answer these practical education questions, the AERI research team will step back to observe and document which formal and informal coordination channels the ministries actually use to absorb and apply this evidence into policy.

Across East, West and Southern Africa, the contexts are distinct and the histories are different. But the institutional dynamics are strikingly consistent: the primacy of the policy window, the gap between formal and informal coordination, the invisibility of internal evidence use, and the political dimensions of evidence uptake. Understanding these dynamics — and supporting governments to build coordination infrastructure that works within and strengthens existing systems — is the work ahead.

By shifting our approach from bringing evidence to co-designing evidence systems, we hope to build a sustainable model for how researchers and African governments can work together to ensure every child builds a strong foundation for learning.

The governments we met in Kenya, Malawi, and Senegal are ready for that work. So are we.

Proud to collaborate with global partners: