Why Evidence Fails to Influence Policy — Even When It Exists

I have seen reports presented in the morning and ignored by the afternoon. New research reports are continuously published. Slides are shared across conferences. Findings are praised in meetings. And then, quietly, everyone moves on. Decisions are made anyway, under pressure, politics, and budget constraints, while the evidence meant to guide those decisions fades into the background.

This experience was a powerful reminder, one reinforced during my recent MIT training on academic engagement in public policy, that the problem is rarely a lack of evidence, but how systems are designed to use it.

As someone who works closely with different institutions on climate and governance policy, I have seen this pattern repeat itself. Somalia is not short on research. Each year, assessments are produced on climate change, food security, education, governance, and economic reform. Many are rigorous. Many are written with genuine commitment. Yet too often they end up in inboxes or shelfs, far from the rooms where real trade-offs are negotiated.

The problem is not a lack of evidence. It is how we design systems to use it. This gap between evidence and action is not uniquely Somalia. Even in highly resourced systems, research struggles to shape policy. Policy scholar Henry Rome has shown that in the United States, vast university research output often fails to influence public decisions because incentives, timelines, and communication channels are misaligned. Researchers are rewarded for journal publications; policymakers are rewarded for managing politics and delivering results under pressure.

Evidence does not fail because it is wrong. It fails because it is not built to travel. The Overseas Development Institute’s RAPID framework makes this clear: policy decisions are shaped as much by political context, relationships, and timing as by data itself. That is why institutions such as the World Health Organization created the Evidence-Informed Policy Network (EVIPNet). Even in health, where evidence can be life-saving, research was routinely ignored unless translated into concise policy briefs and embedded in structured dialogue processes.

To be sure, this is not a failure of intelligence or commitment. Ministers and civil servants operate in environments of urgency and scarcity. No senior official has time to digest a 70-page technical report during a crisis. But acknowledging these realities is precisely why reform must focus on design rather than blame.

Over the past decade, Somalia has made important progress. The Somali National Bureau of Statistics, established under the 2020 Statistics Law, now produces national surveys that did not exist a decade ago. Data infrastructure is improving. But data alone does not change outcomes. When research is used to justify decisions already made rather than to explore alternatives, evidence becomes decoration, not direction.

This pattern is visible in climate policy. Somalia has no shortage of climate risk assessments. What is often missing is systematic learning from how communities already adapt: pastoralists adjusting migration routes, farmers experimenting with crops, and local authorities negotiating access to water and land. Global discussions on locally led adaptation increasingly recognize that solutions detached from political economy and lived experience struggle to endure.

Policy is not a machine into which data can simply be inserted. It is a negotiated, human process shaped by power, incentives, and trust. If Somalia, and Africa more broadly, wants evidence to matter, we must rethink how it is produced and used. First, evidence must begin with real questions, those policymakers and communities are actually grappling with. Research designed in isolation from decision cycles rarely lands at the right moment. Second, translation is essential. A 70-page report has value. But a two-page brief, delivered at the right committee hearing, can change outcomes. Evidence travels through relationships: trusted messengers, repeated engagement, working groups, and dialogue, not through one-off publications. Third, policy must be treated as a learning process. In fragile settings, waiting for perfect evidence is unrealistic. What works better is iterative experimentation, test, learn, adapt. This requires leaders to acknowledge uncertainty rather than conceal it.

Parliamentary committees can institutionalise short evidence hearings linked to live debates. Ministries can require major policy proposals to answer three questions: What do we know? What don’t we know? What will we learn next? Universities and think tanks can establish lean knowledge-translation units focused not on bureaucracy but on briefs and dialogue. Governments and Donors can invest in local evidence ecosystems, statistics offices, think tanks, and universities, as long-term public goods. Evidence will never replace politics. Nor should it. But it can make political choices wiser, fairer, and more grounded in reality. In a country facing overlapping climate, economic, and governance pressures, unused evidence is not neutral. It is a missed opportunity. The real question is not whether Somalia has enough data. It is whether leaders are willing to design systems that turn knowledge into action.