It’s time to experiment with how Europe funds science
An experimental metascience unit would allow the Commission to test innovative ways of awarding research funds with immediate impact

This viewpoint was first published in Science|Business on 14 April 2026.
In the next few months, the EU will set the shape of its next Framework Programme for research and innovation. The institutions are arguing about the size of the budget and what to spend it on, but not enough people are asking a more basic question: how do we learn to spend it effectively?
Some governments already run experiments on their funding systems and are seeing results. They are drawing on metascience, the systematic study of how research is funded and whether those systems actually work. The EU should be doing the same. It needs an experimental unit to start putting its funding processes to the test.
Does peer review actually identify the best research? Do larger grants produce more impact per euro than smaller ones? Would shorter applications produce worse funding decisions? These are empirical questions, and we do not know the answers.
The EU spends tens of billions of euros on public research using processes designed decades ago and refined by political negotiation, not evidence about what works. Establishing an experimental metascience unit in the European Commission, as the Heitor Report recommended, would solve this problem.
What the unit would do
An experimental unit would not be a think tank. It would not merely study how research is funded. It would run programmes. The Heitor Report was explicit about this: the unit needs “maximum flexibility from Commission rules and procedures to be able to rapidly experiment with diverse procedures.” In practice, that means a small team of highly competent people, working not as a separate bureaucracy but through the existing arms of the Framework Programme, such as the European Research Council (ERC), the European Innovation Council (EIC), the collaborative grants of Pillar 2, and the Marie Skłodowska-Curie actions.
The idea is that the unit designs funding experiments, then the host programme runs the calls. This solves a vital problem of political economy: programmes want budget, and the unit brings it. Without external funding, programmes have little incentive to experiment, since their stakeholders depend on the current structure and level of funding.
What kinds of experiments would be run? We can start by learning from other science funders. The UK Metascience Unit, established in 2023, trialled distributed peer review that cut assessment times by 53-65% while maintaining quality. Germany’s Volkswagen Stiftung has experimented with funding lotteries and distributed peer review since 2017. New Zealand and Switzerland have run similar trials.
An EU experimental unit could test programme management based on the US Advanced Research Projects Agencies with genuinely empowered programme directors, innovation prizes for defined societal challenges, or entirely new ways of identifying the most promising researchers.
Ultimately, a metascience unit would be a simplifying force within the Commission. The status quo has seen layers of process accumulate, sometimes out of touch with the outcomes they are intended to achieve. A metascience unit would apply a magnifying glass to these norms, questioning them and reducing bureaucracy in a scientific manner rather than a dogmatic chain sawing of regulations.
The Heitor Report calls for the metascience unit to have a “significant” budget. Even 2% of the expected Framework Programme budget of at least €175 billion would give it substantial room to operate, and that money would not be spent on administration. It would flow through existing programmes as experimental funding, testing new approaches alongside current ones.
Given that the last Framework Programme, Horizon 2020, generated an estimated €429 billion in GDP from approximately €80 billion in investment, even small improvements in how funding is allocated would pay for themselves many times over.
Time to act
Neither the Commission’s draft proposal nor the European Parliament’s early positions have taken up the idea of creating a metascience unit. They should. The legislation is entering the next stage and, precisely because this is not a big-ticket item, it can still be introduced.
The experimental unit seems to have broad support among researchers and reformers, but it keeps getting lost in negotiations. It is not a big enough line item to attract champions and incumbents naturally push for more money in their own programmes. The ERC community pushes for more ERC funding. The EIC community pushes for more EIC funding. National governments focus on what they can get back. Nobody’s job depends on testing whether the system itself could work better.
The EU is about to commit billions of euros to research. We should test whether the processes that allocate it actually work, and where necessary improve them. The legislation is open now, and it will not be open again for seven years. Adding an experimental metascience unit is not a radical proposition. It is the minimum. The EU should act before we fall behind.
Start the unit. Fund experiments. Scale-up what works. That is all it takes.