(Science|Business) ERC evaluators stoical in face of rising workload

For the original publication, please click here.

Academics reviewing applications for European Research Council (ERC) funding are feeling the strain of a rising workload, but are not yet at breaking point, according to feedback received by Science|Business. 

ERC proposals are evaluated by 91 peer review panels, which meet twice a year to shortlist applications that will be evaluated in depth, then to interview applicants and decide on the final ranking. Each panel handles between 50 and 150 proposals on average. While some see lower volumes, others have seen the workload top 250 applications in recent rounds. For example, success rates have fallen from 14.8% in 2023 to 12.2% in 2025 for applications to ERC starting grants, and from 14.5% to 11.2% for ERC consolidator grants.

“It was hard work and we had a lot of proposals to triage, around 20-25 each at the early stage, and then maybe eight to read in greater depth in the second round,” said Laura Machesky, professor of biochemistry at the University of Cambridge, who was a review panel member for three years. “I found this manageable,” she went on. “I was willing to work on weekends and in the evenings to complete my assignments.”

The workload is “high” but “doable,” according to Harro van Lente, professor of science and technology studies at the Faculty of Arts and Social Sciences of Maastricht University, who has been an ERC evaluator in 2022 and 2024. 

Evaluation meetings take place in Brussels, where the ERC Executive Agency supports reviewers as they work together until all evaluations are finished. “Of course, when one returns, other work has piled up, but it is nice to return without further homework,” Van Lente added.

For Göran Hansson, professor of experimental cardiovascular research at the Karolinska Institutet in Stockholm and a panel member in 2023 and 2025, the process involved “quite a lot of work,” but it was “well organised by the staff and, importantly, necessary for a fair evaluation of the scientific merits of all proposals.”

Science|Business contacted 60 evaluators for comment and received six responses.

The rise in application volumes is often attributed to generative AI making it easier for researchers to draft new proposals or repurpose bids originally made to other research funders. Taking this into account also makes it hard to streamline the evaluation process, for instance by dropping the second-round interviews.

“These are notoriously nerve-wrecking for the candidates,” Van Lente said, “while they might also bias the evaluation, as we know from psychological studies on, say, job interviews.” But on the other hand, “these interviews are also to certify that it is indeed the candidate who wrote and defends the proposal, and not a committee at their home university or AI.”

And AI might not be the only factor behind the rise in ERC applications; cuts to national funding also play a part. 

In the UK, for example, public research funding has been redirected from basic research to innovation and government priorities such as AI and defence. “If I want to apply for a grant now to do basic research, I have fewer options than before,” Machesky said. “It is my impression that the ERC remains one of the few major funders now of basic scientific research.”

Managing the flood

The ERC is well aware of the challenge it faces, but is struggling to find an acceptable response. 

In April 2026, it brought in rules restricting resubmissions, but quickly reversed the decision after a significant backlash from the scientific community. “The ERC will now place greater emphasis on other measures to reduce the workload on panel members,” it told Science|Business. These measures will be presented in July, following the final adoption of the ERC work programme for 2027.

The European Commission, which oversees the ERC’s parent programme, Horizon Europe, has proposed moving from the single-submission, two-step evaluation process operating at present to a two-stage application process. The first submission would involve a short proposal summarising the project concept, with only successful applicants invited to submit a comprehensive, evaluator-ready proposal detailing the work packages and budget breakdowns. 

This way, evaluators would only have to assess short outline applications, while the researchers who are unsuccessful at the first stage would be spared the effort of submitting a fully fleshed-out proposal. This might also be an opportunity to ask current ERC grant holders to help pre-filter first-stage proposals. 

As funding decisions often come down to marginal scoring differences, the Commission is also considering introducing randomisation: all proposals meeting a quality threshold would be entered into a lottery, with winners chosen at random rather than purely through peer review, in order to give equally strong projects a fairer chance in competitive calls.

The ERC has already increased the size of evaluation panels, and constraints remain on the number of submissions that can be discussed in their one-week sessions. But one former panel member, who spoke on condition of anonymity, thought that more could be done. ”Some panels seem to be inundated with proposals and could potentially be split in two, requiring additional panellists, of course,” they said. 

In the meantime, the ERC and its scientific council are contemplating the use of AI-assisted processes, triage mechanisms and tightened resubmission restrictions. “Some of these options could be considered in the context of the next Framework Programme,” the agency said.