Blog
The time to be experimental is NOW: Moving from aspiration to action in 2026
6 February 2026
The recently published European Commission Policy Brief on experimentation in research and innovation (R&I) policymaking makes one thing clear: in an era of AI-driven shifts and climate urgency, the traditional, rigid policy cycle is no longer fit for purpose. We need agility. We need rigour. We need policy experimentation.
We developed the brief, A New Approach for Impactful R&I Policymaking, as a strategic call to arms for policy experimentation to become the new standard for closing the continent’s innovation gap. In it we propose that the Commission focus on three main recommendations:
- Building the capacity to experiment through shared literacy and skills.
- Creating opportunities through institutional support such as dedicated units and flexible funding.
- Strengthening motivation and incentives to recognise constructive failures as learning opportunities.
And while the European Commission is signaling a new era of aspiration, great opportunities for action arise in national agencies and across European agencies. As we look toward 2026, the question is no longer if governments should experiment, but how they can embed this scientific approach into the very DNA of public institutions.
No more excuses
There are many reasons, or excuses, for staying in the safe lane of traditional policymaking instead of building experimental policy institutions.
One major reason is the fear of public or political backlash. But that excuse is crumbling under the weight of new evidence. We often assume the public views policy experiments as risky or unfair. Recent findings tell a different story. In a large-scale experimental survey across six EU countries, referencing “experiments” or “random selection” did not reduce public support for policy initiatives. In fact, citizens see experimentation as a pragmatic and responsible way to ensure public funds are being used effectively rather than being wasted on untested assumptions.
Another common hesitation is the fear of an experiment revealing a policy is ineffective. The hard truth is that well-intended policies fail all the time, we just don’t usually measure them. For example, a 2017 study showed that providing export information to businesses actually made those who didn’t less likely to export. Without experimentation, we continue to fund programmes based on intuition that can backfire.Furthermore, the Heitor expert group has called for the establishment of experimental units to test and scale R&I programs. We are moving from a world where experimentation was a nice to have to one where it is a competitive necessity.
Five pillars for experimentation
If 2026 is to be the year of rapid progress towards experimentation being an integral part of decision making, national agencies must move beyond isolated pilots and toward integrated learning systems. Here is how:
1. Give experimentation agency and budget: Experimentation cannot be a side project. To thrive, it needs a dedicated home. Whether it’s a Metascience Unit or an Experimental Unit or Lab, agencies need senior champions with the budget and the legal mandate to apply experimental approaches to the policy cycle. To be truly effective, these units must lead the charge in running iterative rounds of testing, ensuring that learning isn’t just a byproduct of policy, but its structural core. Their mandate should extend beyond trials to the very foundation of evidence: improving the quality and accessibility of data. By turning data into a strategic asset, these units empower agencies to design, execute, and evaluate new ideas with a level of precision that traditional intuition-based policymaking simply cannot match.
2. Kill the ‘fear of failure’ culture: Move from a culture of compliance to one of learning. This requires important internal mindset shifts, including normalising constructive failure: rewarding those who spot the flaws in a plan before it scales. Creating dedicated spaces for policy ideation and experimentation, and incentivising officials to prioritise evidence-informed decisions over simple delivery.
3. Rethink the researcher’s role: For too long, governments have treated researchers as external sources for occasional input. We need to bridge the gap by giving researchers the ability to generate evidence through policy delivery. By embedding researchers within the design phase, we ensure that findings aren’t just academic curiosities, but actionable insights capable of driving both the design and delivery at scale of impactful policies.
4. Build momentum: Not every experiment needs to be a multi-year large-scale impact evaluation. We should embrace small gear experimentalism: low cost and rapid trials like A/B testing or experimental surveys that build institutional confidence about the method, while generating relevant impact and insights. These quick wins can prove that data-driven decisions are faster and safer than operating on assumptions.
5. Create blended, flexible funding: The rigid structures of traditional funding, including many Horizon Europe frameworks, often stifle the very experimentation they aim to promote. We need dedicated Experimentation Funds that blend resources for both implementation and testing across a portfolio of solutions. These funds must offer the radical flexibility to pivot, or even stop projects entirely, based on early evidence. Innovation is rarely linear; it is time our funding reflects that reality. By funding the learning journey, we ensure public money follows actual impact rather than remaining stuck to an original roadmap that leads nowhere.
The first step is the hardest
Our recent policymakers survey shows a clear trend: the first experiment is always the most difficult. However, those who try it almost universally find the process valuable and would do it again.
The European Commission has set the stage. It’s time for R&I agencies to stop preparing to experiment and start doing it.