Over the past two decades, the field of evidence-informed policy (EIP) has matured significantly. We’ve got more data, more researchers, more think tanks, and more donors funding better research uptake. And yet, policy remains stubbornly “stuck” in many places. It’s not because we lack solutions. It’s because we’ve misjudged the problem.
If we want evidence to cut through and support real policy change, we need to stop thinking of evidence as the answer—and start thinking of it as a contributor to a much more complex, political, institutional, and human conversation.
The real challenge: policy is political, not just technical
The evidence community still spends much of its energy asking: “What works?” But this question alone has limited utility. As the discussions at the 2025 OTT Conference suggest, the real question is: Why doesn’t what we know already work?
The answer lies in the messy and often uncomfortable terrain of politics, institutions, incentives, and trust. Opportunities to influence policy exist, from agenda setting to implementation and evaluation, but each of these spaces is highly contested and rarely linear. Competing interests, implicit trade-offs, and conflicting values dominate decision-making spaces. Evidence, in this context, is just one of many inputs, and not always the most important one.
Policy reform has a tendency to stall. Despite vibrant civil society, independent media, and formal oversight mechanisms, thorny issues persist. From unresolved ideological debates about economic ownership to the equitable delivery of public services, the same challenges arise repeatedly—not because we lack evidence, but because the political will, institutional structures, and social foundations necessary to act on that evidence are absent or misaligned.
“Thorny issues”: The barriers we’d rather not talk about
A helpful concept emerging from recent work by OTT is that of thorny issues – those complex, politically sensitive, and often deeply embedded challenges that derail evidence use and resist easy resolution. They can be grouped into two categories:
1. Contextual thorny issues: rooted in a country’s political and social fabric:
- Corruption and state capture distort the use of evidence and reduce transparency.
- Political polarisation, which turns evidence into ammunition rather than insight.
- Weak institutions with limited capacity for research use or independent analysis.
- Social norms and distrust, which affect how evidence is received, especially if it comes from perceived outsiders or elites.
- Vested interests, such as corporate lobbies, that actively undermine or suppress inconvenient truths.
2. Sector-specific EIP challenges: internal to the evidence field:
- Flawed business models, where research is shaped more by donor expectations than policy relevance.
- Incentive misalignment, with evaluations designed to satisfy funders, not end users.
- Opaque funding, which can raise suspicions about independence and neutrality.
- Weak incentives for communication, leaving good research buried in unread PDFs.
These are not abstract problems. They’re real, and they affect whether evidence makes it into policy or not.
Strategies that actually work (even if they are not easy)
If the barriers are this complex, then our responses need to be equally sophisticated. Here are some ways think tanks and evidence advocates can work smarter—and with greater humility. Damien King, Tom Hashemin, Rakesh Rajani, Sandy Africa, and the participants of a session on “getting policy unstuck” at the 2025 OTT Conference offered the following recommendations:
- Stop selling technical solutions to political problems: Recognise that most policy issues are political and institutional at heart. If you frame your work purely in technical terms, it will be ignored—or worse, used against you.
- Build trust—it’s the currency of change: Evidence matters. But in most political environments, trust matters more. Build relationships early. Co-analyse problems. Co-create solutions. That’s how you get ideas adopted.
- Start with their agenda, not yours: To effectively influence someone, you must understand them. That means approaching policymakers with curiosity and empathy. What are their priorities? What keeps them up at night? What football team do they support? The more you start with their world, the more likely they are to listen to yours.
- Create political opportunities, not just proposals: Most policy recommendations fail not because they’re bad ideas, but because there’s no demand for them. Don’t push a solution. First, create a space where that solution makes political sense. Sometimes that means using public pressure or media attention to create urgency – think sewage spills, not spreadsheets.
- Be useful: help them do less, better: Governments are overwhelmed. Don’t add to their workload. Instead, frame your work as making things simpler, more effective, or more politically attractive.
- Get serious about communications: Research is relatively easy. Embedding ideas in people’s minds is much harder. Every think tank should invest in strategic communications 0- even a small, dedicated team can make a big difference. Think about audience, channels, timing, and tone. And be ready to repeat your message, again and again.
- Understand how the government really works: This is where bureaucratic ethnography comes in. Map out how decisions are made. Who holds informal power? How does information flow (or not)? Civil servants matter – they’re often the ones who will make or break your proposal. Engage them early. Respect their constraints. Speak their language.
- Give policymakers ownership: Let them take credit. Let them reshape your ideas in ways that make political sense. Most politicians and bureaucrats don’t want to be told what to do. They want support, not instructions.
- Work across the system: Policy change doesn’t stop at the minister’s desk. Engage civil society, the media, technocrats, and citizens. Mobilise constituencies around your ideas. Build ecosystems, not just briefing papers.
- Learn from others: There is a wealth of knowledge about how to navigate complex political environments. These lessons are often undocumented, overlooked, or undervalued by think tanks.
Moving from evidence to influence
Getting policy unstuck requires more than better research. It demands a shift in mindset – from evidence generation to systems thinking, from neutrality to political awareness, and from outputs to relationships.
The field of evidence-informed policy must evolve. It must face the thorny issues head-on, invest in strategic engagement, and listen more closely to the people it seeks to influence. This includes politicians and bureaucrats, as well as civil society actors, journalists, and citizens. And crucially, it means learning from those who’ve had to do this work in the most difficult contexts – often without fanfare, and always without guarantees.
In the end, the most impactful think tanks are not those with the best data, but those with the deepest understanding of how power works, the humility to listen, and the courage to speak when it matters.