The binary divide between old and new universities persists in the higher education sector, official figures published this week confirm.
The statistics highlight the uphill struggle facing research universities trying to open their doors to working-class students, and the continuing problems for new universities aiming to reduce student dropout rates.
Performance measures showing how UK universities fared at attracting students across social classes, retaining undergraduates and generating research income were published by the Higher Education Statistics Agency this week. The 2002-03 figures show a familiar pattern of major research universities falling short of their targets for recruiting working-class undergraduates.
Conversely, some new institutions continue to have dropout rates of more than 25 per cent - with a high of 37 per cent at Napier University - while some research universities retain 90-98 per cent of undergraduates.
Paul Temple of the Institute of Education, University of London, said:
"There are two higher education systems depicted in these statistics: research-intensive universities, recruiting well-qualified students, almost all of whom graduate; and teaching-led universities, doing little research, recruiting less qualified students, many of whom struggle to complete their courses or remain in higher education.
"The system is putting pressure on the few transitional institutions, those towards the bottom of the pre-92 (university) league tables: it seems likely that, on present trends, they will end up in one group or the other."
Sir Michael Sterling, chairman of the Russell Group, said the proportion of state-school pupils attending the group's 19 institutions was rising faster than in the higher-education sector overall.
The percentage of undergraduates from state schools nationally rose from 86 per cent to 87 per cent between 2001-02 and 2002-03, while the proportion of state-school pupils at Russell Group institutions rose from 72 per cent to 74.5 per cent.
"This is clear evidence that the Russell Group is taking the issue seriously," Sir Michael said. "But, of course, we are all determined to maintain high standards, that is what the group stands for. We are not about to try to achieve things by manipulating the entry to take weak students rather than strong students."
But Michael Driscoll, chairman of Campaigning Modern Universities, the umbrella body representing post-92 universities, said there was a "fairly obvious link" between participation and dropout rates.
"In response to the Government's desire to see more students coming from backgrounds where there is less of a tradition of involvement, we take more risks, and the dropout rates are a reflection of that," he said.
Professor Driscoll called for the cost of widening participation to be fully funded by the Higher Education Funding Council for England and said the current premium on funding per head paid to universities to support widening access should be raised.
He added that universities were unjustly "penalised" by losing funding if a student dropped out midway through a course, leaving institutions to carry the cost of any teaching that had taken place.
"It's not a failing of the university, it's a structural problem to do with the composition of the student body in different universities," he said.
The figures also show that the dropout rate for mature students is higher than for full-time undergraduates aged 20 or less. At 70 per cent of UK universities, the dropout rate for younger entrants is less than one in ten.
Dr Temple said that the overall dropout rate in the UK remained among the best for members of the Organisation for Economic Cooperation and Development, with only Japan performing significantly better.
"Of course, the overall UK figure hides big differences, relating strongly to the entry qualifications of students," he said.
"Non-continuation is strongly associated with qualifications on entry to higher education - there is an almost straight correlation showing that good predictions of success can be made on the basis of A-level scores. w "The clear policy implication is that the problem of non-completion needs to be tackled in schools or colleges - it's too late by the time people get to university."
The Hesa figures show that 825 young people embarked on higher education courses in 2002 - across the arts, humanities and sciences - without having any previous qualifications.
The data also reveal that more than half (51.1 per cent) of Wolverhampton University's undergraduate entrants aged under 20 were from lower social class groups, far more than its 37.3 per cent target for widening access.
Meanwhile, research universities such as Warwick, York, Durham, Exeter, Bristol, Oxford and Cambridge all missed their widening participation targets among their under-20s intake from the lower social groups by between 3 and 8.5 per cent.
But York, Warwick, Oxford and Cambridge were among those offering the best ratio of research income to staffing costs.
Overall, the percentage of young people from low participation areas who embarked on a university education rose from 12.3 per cent to 13.3 per cent between 1998 and 2002. But participation in areas of England with a poor record of sending young people to higher education lagged behind both that of Scotland and Wales.
By 2002, 18.6 per cent of young people from low participation areas of Scotland were entering higher education and 16 per cent of their counterparts in Wales, compared with 12.5 per cent of young people in England and 10.1 per cent in Northern Ireland.
Liz Allen of lecturers' union Natfhe said the statistics offered only a "partial picture" of the state of higher education and gave no insight into the underlying trends.
"You can see that it is the post-92 (institutions) that do a major job in recruiting students from lower social classes and low-participation neighbourhoods, part-time and mature students," she said.
"If you look at their performance against the benchmarks they are doing well at retaining students, bearing in mind these are set to reflect the reality - that students with lower entry requirements will find it harder."
Ms Allen added: "Inevitably, if you broaden the net, you will take in students who may find it harder to stay in higher education for a range of reasons." Full tables
Dropout rates among full-time students 2001/02
Highest/lowest participation rates form lower social class groups
Research income generated per academic costs