The British system of attribution - whereby every pound a government department spends on the European Framework Programme for science is subtracted from its research and development budget - needs reform, the Royal Society meeting "Science Funding: the European Dimension" heard this week.
Any rise in UK contributions to European science projects means money subtracted from national schemes. The majority of science funded through the Framework programme is applied in nature.
"Many people feel that basic science ought also to be funded in Europe," said Sir Tom Blundell, professor of biochemistry at the University of Cambridge and a former head of the Biotechnology and Biological Science Research Council. But there was "a total disincentive" to move from applied areas of research to basic areas.
The Ministry of Agriculture, Fisheries and Food and the Department of Trade and Industry lose most because of attribution. But if European programmes emphasised basic research, it would be the Office of Science and Technology and the research councils that would be worst hit and the UK could lose much control over which basic research it funds.
Howard Newby, vice-chancellor of Southampton University, said that "almost all universities would be bankrupted" if they won too many European contracts. The EU pays such low overheads on its work that universities cannot afford to be too successful in bids.
The government's chief scientific adviser, Sir Robert May, criticised the way the research assessment exercise allocates funds to university departments, as "looniness". The UK tops the international league table for collaboration, he said, with half the papers in science, medicine and engineering having three or more authors from two or more institutions.
Sir Tom Blundell said that in his field papers resulting from international collaborations were cited nearly twice as often as non-collaborative papers.
Leader, page 11; audio coverage website http://www.thesis.co.uk