"Institutionalised sexism" is denying female academics recognition for the research activity that is vital to their career prospects, a study by the Association of University Teachers has claimed.
The AUT found that men in some universities are up to five times more likely to be classified as "research-active" and entered in the research assessment exercise compared with female colleagues.
"These findings are shocking," an AUT spokesman said. "There is clear discrimination in higher education. The RAE should be smoothing this out, instead, it is magnifying the problems."
The study by Stephen Court, the AUT's senior research officer, found that across the sector, among academics engaged in both teaching and research activities, males were 1.6 times more likely than their female colleagues to be counted as "research-active" and to have had their work submitted in the 2001 RAE.
At the University of Wales College of Medicine, the odds for men were 4.9 times higher, and at Wolverhampton and Plymouth universities men were more than three times more likely than women to be research-active.
Just nine of the 117 institutions in the study offered women equal research opportunities.
The AUT said that women were being excluded despite the funding council's efforts to ensure that staff were not penalised for taking maternity leave or other career breaks. "The Government should be deeply worried about this institutionalised sexism," it said.
Sally Hunt, AUT general secretary, said: "It is time the buck stopped being passed around the system - departments, institutions, the funding councils and the Government. It is time one of them took responsibility for dealing with this."
The report, Gender and Research Activity in the 2001 RAE , found that at the time of the RAE, 39 per cent of 146,000 academic staff in the UK were women. But of the 43,000 staff recorded as research-active, a quarter (25.2 per cent) were women.
Nineteen per cent of all female academics were included in the RAE, compared with 37 per cent of men. "Male staff were therefore 1.9 times more likely than their female colleagues to be counted as research active," the report says.
Figures relating to staff whose employment was categorised as "teaching and research" - to factor out teaching-only staff and those exclusively on research contracts - show men to be 1.6 times more likely than women to be research-active.
The study found that research activity for women declined mid-career, supporting concerns that taking a break harms prospects. For female teaching and research staff, 36.7 per cent of those aged 35-39 were research-active, but this fell to 33.2 per cent among those aged 40-44 and to 30.2 per cent for those aged 50-54.
The biggest gender gap, the report says, is at new universities, higher education colleges and specialist institutions, particularly medical schools.
Roland Levinsky, vice-chancellor of Plymouth, said "the whole structure of the university has changed dramatically" since the 2001 RAE. Plymouth estimates that the proportion of its research-active female staff is now per cent.
A spokesman for the Higher Education Funding Council for England, which runs the RAE, said: "We take equality issues very seriously and have done so in relation to the 2008 RAE." The first published guidance on the next RAE states that equal opportunities will be "a fundamental principle" underlying its design and conduct.
Hefce said its own analysis showed that among eligible staff, 74 per cent of men and 57 per cent of women had been submitted to the 2001 RAE. After taking account of factors such as the different subject profiles of men and women, it found no significant difference between men and women for academics aged up to 30 and those aged over 50, although there was an unexplained difference for the age range 31-49.
"We are considering ways to explore the reasons for these remaining unexplained differences," he said.
· A distinguished scientist should be recruited to help ministers "challenge the conventional wisdom" of Whitehall, says the Butler inquiry into intelligence and the Iraq War. This week's report calls for the Government to consider recruiting a scientist "to undertake a part-time role as adviser to the Cabinet Office".
THE TOP TEN RAE GENDER OFFENDERS
% of female academics
counted in 2001 RAE Female to male odds of
being counted in 2001 RAE University of Wales College of Medicine 9 1:4.9 Wolverhampton University 5.3 1:3.3 Plymouth University 7.9 1:3.2 Edge Hill College 5.4 1:2.7 Bath Spa University College 13 1:2.7 Robert Gordon University 9.7 1:2.6 Salford University 11.1 1:2.6 Teesside University 5.9 1:2.5 St George's Hospital Medical School 17.1 1:2.4 Buckinghamshire Chilterns University College 10.9 1:2.3
The figures include only staff categorised as being employed in teaching and research
'Subjects like gender studies are judged inferior'
Gillian Howie believes the type of research women often do is not taken seriously by the male-dominated research establishment.
Dr Howie, a senior lecturer at Liverpool University's department of philosophy, said male heads of department, university managers and leading academics judged subjects such as gender studies and feminist theory as inferior and peripheral.
"Sometimes I think we overplay the impact of families because most women keep producing the work despite these pressures," said Dr Howie, the director of the Institute of Feminist Theory and Research and a mother of two. "The problem is getting the work itself properly recognised."
She was considering whether the gender-specific elements of her work should be submitted for the 2008 research assessment exercise.
She said: "There is the risk it will not be seen as appropriate." She added that a colleague was told by her line manager that her work would not be submitted - before he had read it. "He did change his mind, but it shows the entrenched attitudes women face."