A year that tested us — and clarified why our work matters
As this year draws to a close, it is tempting to focus on achievements and milestones. But 2025 has not been an easy year for the development sector, or for many of the partners we work with. It has been a year of abrupt funding losses, rising needs, and deep uncertainty — one in which difficult decisions had to be made under pressure, often with fewer resources and less room for error.
For many governments, funders, and implementing organizations, the challenge has not been a lack of ambition, but a lack of slack: time, money, and certainty have all been in short supply. In that context, the question we have heard most often is not “What else can we do?” but “How do we make better decisions with what we have?”
That question goes to the heart of why 3ie exists.
Evidence as an ethical responsibility, not a technical luxury At a time when resources are shrinking and needs are growing, using evidence well is not simply about efficiency. It is about responsibility. Decisions taken today — about social protection, climate adaptation, education, peacebuilding, or livelihoods — shape real lives, often irreversibly. When evidence is ignored, misused, or oversimplified, it is rarely decision-makers who bear the costs.
This year has reinforced a lesson we have learned repeatedly: knowing what works is not enough. Decision-makers also need to know how well something works, for whom, at what cost, and under what conditions — and they need that information when decisions are actually being made, not months or years later.
From producing evidence to strengthening evidence cultures One of the most important shifts in our work this year has been a deeper focus on institutional evidence cultures.Through the Global Evidence Commitment (GEC), we work with development agencies and funders to strengthen the incentives, processes, leadership signals, and skills that make evidence use possible in practice.
The GEC is deliberately modest in form but ambitious in intent. It recognizes that evidence use does not fail because people are unwilling to learn, but because institutions often make learning difficult. Progress, therefore, lies not in one-off reforms, but in steady changes to how decisions are prepared, questioned, and revisited over time — especially in periods of volatility.
Strengthening evidence cultures is equally critical inside governments, where many of the most consequential decisions about public spending and reform are taken under intense political and fiscal pressure. Supporting governments to use evidence systematically requires long-term engagement, trust, and a realistic understanding of institutional constraints.
This year, we took an important step in that direction with the launch of the Egyptian Observatory for Evidence-Based Policymaking (EOEPM), in partnership with government institutions and local think tanks. The Observatory is not a research project or a one-off advisory exercise. It is an effort to embed evidence more deeply into how policy questions are framed, how options are assessed, and how learning feeds back into future decisions — including through rapid but rigorous synthesis that responds to real-time policy needs.
We have also continued to invest in long-standing partnerships in West Africa through the WACIE programme, focused on strengthening evaluation and evidence-use capacity within government and public health institutions. This work is deliberately practical: building skills, peer networks, and institutional routines that make evidence use more feasible over time.
In India, our collaboration with the Government of India — particularly through large-scale livelihood and social protection programmes — continues to show what sustained engagement can achieve. By embedding evidence generation, synthesis, and learning within programme cycles, this work has helped decision-makers assess not only whether interventions work, but how they can be improved, adapted, or scaled. Across these contexts, a common lesson stands out: evidence cultures grow when governments are treated not as recipients of findings, but as co-producers of learning.
Making evidence usable at speed: the Development Evidence Portal and AI
The Development Evidence Portal (DEP) remains a central part of our contribution to the global evidence ecosystem. But access alone is not enough. Policymakers today face severe time constraints, information overload, and heightened uncertainty. The challenge is not just finding evidence, but making sense of it quickly and responsibly.
That is why we are investing in enhancing the DEP with AI-supported tools that improve search, navigation, and summarization across a large and complex evidence base. These tools are designed to help users surface relevant evidence more efficiently, compare findings across contexts, and understand both what is known and where important uncertainties remain.
This work is guided by a clear principle: speed must not come at the expense of rigour or accountability. In an era where AI-generated content can easily overstate certainty or obscure nuance, we see a responsibility to model careful, transparent, and evidence-informed use of these technologies. The goal is not automated answers, but better questions — and better-informed decisions — at moments when they matter most.
What evidence is telling us — and why humility matters Across our work this year — including in peacebuilding, livelihoods, and social protection — evidence has continued to challenge assumptions. Some approaches that are widely believed to work do not perform as expected; others show promise only under specific conditions. This is not a failure of evidence, but its value.
Recent peacebuilding evaluations, for example, suggest that local, context-specific mechanisms can reduce conflict, but only when they are well designed, adequately supported, and embedded in broader political processes. There are no universal solutions, and pretending otherwise risks doing harm. More broadly, evidence should make us more cautious, not more confident. It should help decision-makers recognise uncertainty, weigh trade-offs, and avoid overclaiming impact — especially when stakes are high.
Looking ahead to 2026 As we look to 2026, we do so with clear eyes. The development landscape will remain uncertain, and the pressure on institutions to justify decisions and demonstrate value will only increase. Our focus in the year ahead will be on deepening the work that matters most: strengthening evidence cultures in organizations and governments, and making high-quality evidence usable under real-world constraints.
We will continue to invest in partnerships that allow for learning over time, adapt our tools and methods — including the responsible use of AI — to better meet decision-makers’ needs, and remain open to revisiting how we work as contexts and constraints evolve. In a period of transition for the sector, steadiness, relevance, and integrity will guide our choices.
With thanks and best wishes
We are deeply grateful to the governments, funders, researchers, and implementing partners who have worked with us through a challenging year, and to the 3ie team whose professionalism and commitment have made this work possible.
As the year comes to a close, we wish you and your families a restful holiday season and a healthy, peaceful start to 2026. We look forward to continuing this work together in the year ahead.
