Forthcoming public spending cuts may force the UK's research councils to rescind some grants and introduce measures to limit the demand for those that remain, a senior figure from the Economic and Social Research Council has warned.
Phil Sooben, director of policy and administration at the ESRC, told the UK Research Conference in London last week that the councils were expecting a 25 per cent decrease in their funding in light of the Comprehensive Spending Review on 20 October.
He emphasised that the ESRC did not have a "secret list" of things it would cut, but said that retrenchment on this scale meant that "the only way to release significant funds for new activities would be to stop existing ones". He said the ESRC would also have to be "more selective" about what it funded.
"You can't shave the same amount off everything," he said. "You have to be more strategic and selective about what you preserve, what you cut significantly and what you cut completely."
Responsive-mode funding, where researchers propose their own projects, was likely to be particularly vulnerable, he said, because the research councils would be under increasing pressure to direct funding towards areas of governmental priority.
"We'll be pushed even harder on the impact agenda and we will have to demonstrate that the work we are going to fund has wider relevance - without throwing out of the window support for blue-skies research," he said.
"Some streams that are currently entirely open may have to be steered to priority areas."
Mr Sooben also warned that if grant application success rates - which already are only about one in five - fell much further, the ESRC may have to find a way to curtail researchers' current practice of submitting as many applications as possible.
"The logic is that if you take enough shots at goal, eventually you will score. But my answer is if you continue to miss you'll be dropped from the team," he said.
He did not commit the ESRC to following the example of the Engineering and Physical Sciences Research Council, which in April began restricting serially unsuccessful applicants to one application during a 12-month "cooling-off period".
"But we'll have to look at that possibility rather more seriously than we have in the past," he added.
Mr Sooben was urged by a number of questioners to protect PhD studentships. He said that strength of feeling was "an important message to take away", but warned that studentships may have to be focused on a smaller number of universities, with more central direction in certain academic areas.
He said he hoped the overall science budget would be unveiled on 20 October, but added that the research councils' allocations would not be known until December. He warned that it was likely that they would not be in a position to finalise their funding plans until early next year, shortly before their implementation in April.
Mr Sooben also warned that whatever their CSR settlements, the research councils were obliged to cut their running costs by a third.
"That figure is pretty drastic and will change the relationship universities have with the councils," he said.
Not every university department should be expected to have the same level of research impact, according to John Scott, pro vice-chancellor of research at the University of Plymouth.
Professor Scott told last week's UK Research Conference that there should be lower impact expectations for departments that are largely engaged in blue-skies work or have a high proportion of young researchers with no track record.
"Blue-skies research has a different relationship (than applied work) to impact," he said. "It is more long term and iterative. At any one time and at any one institution, it is unlikely to be able to demonstrate impact. But there is an intellectual division of labour and not everyone needs to be able to demonstrate impact at every point in time."
The implementation of the research excellence framework, which will be used to allocate about ?1.5 billion in annual quality-related research funding in the UK, has been delayed to allow for greater consideration of plans to measure the impact of work.
It was originally proposed that impact should make up 25 per cent of REF scores.
Professor Scott also criticised the framework for treating impact as a linear process and for failing adequately to acknowledge wider "cultural" effects, such as television appearances by researchers. "We should talk about wider engagement, not just impact," he said, adding that the research councils' emphasis on "pathways" to impact was a "more sophisticated" approach.
Nigel Vincent, vice-president for research and higher education policy at the British Academy, warned that it was "suicidal" to argue that certain subjects, such as those in the humanities, should not have impact.
"It won't be everyone who has it but there are ways to define it and we simply have to," he said.
He added that impact was "not going away", however "lucky or smart" its opponents were.