Sir

During the past two years, European scientists, politicians and science policy-makers have engaged in a great deal of discussion about the desirability of creating a pan-European funding agency for basic research, the European Research Council (ERC) (see, for example, Nature 421, 881; 200310.1038/421881b). The arguments in favour of creating an ERC are persuasive (see http://www.esf.org/newsrelease/63/ERC.pdf), but there remains much debate on how to organize and fund it. We propose a 'learning structure' for the ERC that would allow it to start small and from scratch, with minimal investment in capital and permanent staff, using the proven expertise of existing grant-awarding bodies.

Given the virtually unanimous opinion in favour of creating an ERC, we may optimistically assume that the stakeholders will divert sufficient funds to provide a small budget for an embryonic start-up in, say, 2004–05. The stakeholders will include the European Union (EU), national research councils (countries that currently do not have research councils should be required to establish them if they wish to participate), and perhaps also other European science-funding organizations and charities. To avoid claims of 'juste retour' — from participants wanting to ensure that they get as much out of the scheme as they have put in — countries that are far below the EU average for research funding would simply pay directly for the grants that are awarded to their own researchers through the ERC selection process rather than providing general start-up funds.

A 'learning structure' must be established to administer this start-up budget and to generate the experience necessary to build the 'grand vision' of an ERC. The ERC must have a flexible, responsive and administratively lightweight structure, learning from others' past mistakes as well as from examples of best practice. Peer review of grant applications, for example, must be rigorous and transparent, applications must be processed quickly and awards must be paid on time.

We propose that the people best placed for this role in the embryonic ERC are administrators seconded from the stakeholder funding agencies, especially national research councils. Administrators from appropriate agencies would constitute 'consortia' representing one area of research such as genomics or particle physics. Each consortium would construct a 'study section' of expert scientists to evaluate and select applications for funding, with one administrator as coordinator. A consortium for genomics, for example, might be hosted by the European Molecular Biology Organization (EMBO), where administrators from all nations would benefit from the 'best practice' they observe at EMBO and the networking opportunities afforded by working with a multinational group. They would return to their national agencies after their secondment to the ERC having gained experience that will help to improve their country's science administration.

The administrative consortia would, however, need some central resources to create databases and websites, and the ERC would need a directorate and a headquarters — probably in Brussels, for proximity to EU organizations. The ERC will also need a president who has earned the respect of the community and who is chosen from, and approved by, a selection nominated by the national academies.

Beneath the president there would be two or three boards of governors representing broad branches of science and comprising academic leaders in the appropriate fields as well as policymakers and, where appropriate, representatives from industry. The governors should serve a fixed term — three years, for example — with new governors being nominated and elected by a stakeholders' council made up of representatives of the funding agencies contributing to the budget of the ERC.

This interim structure is flexible and transparent, not simply administering grants, but providing a 'learning mechanism' for the national programmes and governments of less advantaged European nations to improve their domestic research funding and administration structures. In the long term, this would help to level the playing field for research across Europe.

The ideas proposed in this letter are discussed more fully in an article that will appear in the October issue of The ELSO Gazette at http://www.the-elso-gazette.org/magazines/issue16/features/features1.asp