Most universities want one,many are developing one, butopinion on how to run them varies. Graduate schools are in vogue aspostgraduate education grows apace. David Charter examines the opportunities and conflicts confronting administrators.
Put representatives from Britain's nascent graduate schools together in one room and it soon becomes clear they have no common blueprint for organising postgraduate education.
The patterns vary from university-wide schemes, as at Warwick, to groupings of faculties (for example Manchester) and graduate schools based in individual departments (eg Bristol). Some have problems with resources; others believe that there will soon be a big "crunch" as the schools compete with departments for cash.
But the 34 institutions shown by a recent United Kingdom Council for Graduate Education survey to have established graduate schools do concur on one point - their aim is to improve the quality of graduate education. This is particularly important in view of the council's own prediction that postgraduate numbers are set to grow rapidly.
"Many of those who will be taking up graduate courses will be self-financing and will therefore be even more keen than undergraduates to ensure they obtain full value for money," concludes the council.
"They will seek out information about the extent to which the graduate structures of an institution have successfully added value to previous graduate students.
"Those institutions which have employed a graduate school for marketing purposes or as window-dressing will find it increasingly difficult to compete in this market atmosphere."
But that does not necessarily mean the graduate school has to have a separate physical location.
Warwick University established its school in January 1991 but, as its head, Bob Burgess, says: "We decided to have a university-wide model because we are a centrally driven, centrally resourced university. The fundamental building block is the academic department. There is no building labelled Warwick Graduate School. It exists in every element of the university."
He adds: "If graduate students are funded for 46 weeks of the year by research councils, it is no good saying 'I will see you in term time' because term time does not exist. In graduate education we are talking about a 50-week year. The research library and other facilities have to be available for those students."
Mr Burgess does not favour isolated graduate schools which split research-only staff away from teaching-only staff.
"If you are going to start dividing people in that way you are going to start destroying one of the great virtues of the UK system," he says.
"If you are going to preserve good quality graduate education you have to argue that staff are employed to teach undergraduates and postgraduates. The point is, does it make a difference to the graduate student on the ground. You are not going to get a sea change in which the graduate student comes in for a two-hour seminar and suddenly gets to know 2,000 people. But certainly the experience I have had means you overcome isolation and improve facilities."
Students themselves, represented by the National Postgraduate Committee, have welcomed graduate schools because they raise the profile of the postgraduate experience. But NPC general secretary Jamie Darwen says: "As to what they have done so far, I think it is fairly slow progress. There is a worry that in a lot of cases it will be purely superficial and just a marketing device, especially for attracting overseas students."
For now, graduate schools are largely a movement of the old universities. Some 85 per cent are in the pre-1992 universities, although some, like Oxford, have rejected the idea.
Richard Hughes, graduate studies officer and assistant registrar at Oxford University, says: "I think the university reflects many of the concerns about putting an additional structure on top of the existing structure and it is probably more convinced these issues can be addressed in other ways which will not change the character of graduate education in Oxford, which attempts to have a three-fold foundation of college, department and university.
"Until the university is convinced we should move on to something different we will continue to make it work as well as we can."
Leeds rejected the idea of a graduate school in favour of research schools. Professor Peter Scott of the university's centre for policy studies says: "It would be difficult to create a graduate school in Leeds because its departments are cost centres so it is a very sectionalised university."
He believes there is a considerable conflict between graduate schools and traditional departmental structures.
"They are more of a radical idea than many people think. I suspect the crunch is coming soon - if we are going to make graduate schools work we have got to make them resource centres. Those resources have to come from somewhere and it has got to come from departments, which is when the crunch will come."
Derek Parsons, departmental administrator in the department of physics at Bristol University, said its graduate schools were based on departments for coherence. "I think it avoided a lot of problems, such as competing for funds within the faculty or university," he says.
However, Brian Lewis of London Guildhall University sees a clear division between departmental and graduate school roles.
"We recognise there are institution-wide requirements for taught masters programmes and research. We are not moving taught masters programmes out of departments - teaching and supervision is a matter for the departments. There are matters institution-wide, like generic skills, quality controls, marketing and recruitment."
Salford chose a university-wide model when it established its graduate school in April last year.
Dr Kathleen Whyte, senior assistant registrar of the graduate school, says: "In effect it is a faculty of the university, headed by a director, managed by a board similar to a faculty board. A university-wide model was chosen because we are a relatively small university and our main need was for integration and cohesion."
But she says: "Many developments have been impeded by lack of resources. It means developments follow very far behind the ideas."
At Manchester the structure of the graduate school is evolving from the university's six research and graduate schools. The three covering biological science, medicine and science have since combined in the area of postgraduate education but continue to develop research separately.
"Each school should develop a model taking into account its own needs and circumstances, taking into account all the policies of its research councils," says Professor Tony Trinci, pro vice chancellor for research and graduate education. Each school is headed by a graduate dean and there are corresponding undergraduate schools.
There is another way of doing things: groups of universities get together to provide joint postgraduate programmes. But this has been slow to catch on so far, with Scotland the notable exception.
Scotland has a doctoral programme in economics, where eight universities provide staff for a first year based in Glasgow, after which those attaining their MSc can register for a PhD at any of the institutions involved. All through the programme, postgraduates can go on common weekend training schools, conferences and seminars.
Warwick's Bob Burgess says: "I cannot see there is much future in everyone competing to get three or five students on well nigh identical MA courses when you can get something much more powerful drawing on the expertise of a lot more staff. Regional provision is likely to be much more on the agenda."
Co-operation could help save money. Manchester's Tony Trinci says: "One area we all have difficulty maintaining is libraries and journal stocks. In my case I can walk within 15 minutes to three different centres. There has got to be a better way of organising a regional resource."
Some of those involved in establishing graduate schools believe the various ideas about how best to structure them may reflect more fundamental concerns about the developing nature of postgraduate qualifications themselves.
Mr Burgess says: "There has been a tendency to think we are very clear what we mean by graduate education but people are realising there is no road map of what constitutes graduate education in the UK.
"There is no one model which is appropriate. There is a diverse range of models but if we can make sure the graduate voice is heard in the institution and make sure the graduate experience is a good one then graduate schools will have been a good thing for UK higher education."