Cheques and balances
Peter Knight looks at the winners and losers in university teaching
Which university wins and which loses when it comes to funding teaching? Or is the Higher Education Funding Council for England allocation system so fair and equitable that all universities get the same money for the same work?
Despite publicity about the success or failure of individual universities in the research assessment exercise, no attention has been paid to their funding for teaching. The existing system massively rewards certain universities while significantly underfunding others. Have these winners and losers been chosen by design or is this result an accident of history?
Allocations are complex. Universities receive money for teaching both from fees and from the funding council. Excluding initial teacher training, which is transferring to the Teacher Training Agency, there are 15 subject groups, two levels and two different modes of study to be analysed. This means that over all the universities and colleges there is the possibility of 60 different units of funding.
Dividing the money a university gets by its number of students will not give a meaningful result. A university with a predominance of expensive subjects such as science, engineering or medicine, will receive more money per student than a comparable university that is teaching mainly in the humanities and arts.
Unsurprisingly, a university with a predominantly part-time student population is likely to have less money per student than a competitor with mainly full-timers.
A further problem is that the allocation from HEFCE does not represent the actual cost of that subject. It would be quite proper for a university to receive generous funding for one subject and to subsidise a less adequate level of funding in others.
It is possible to devise a system that resolves the ambiguities in HEFCE's allocation of teaching funds and exposes the winners and losers. It is reasonable to expect that universities offering subjects within the same academic subject category at the same level and mode will have comparable costs. All universities that are teaching engineering on full-time, taught courses might be expected to have the same allocation. This is a single one of 60 cells on which the system is based. In fact universities vary considerably in their funding for engineering. One of the most generous allocations goes to Imperial College at Pounds 2,738 per student while one of the least is Bournemouth University at Pounds 1,362 per student. The average for all universities in engineering is Pounds 2,225.
The fact that a university has a generous allocation in engineering does not necessarily mean that it will do well over all other subjects. It might be that although engineering is well funded by HEFCE at that university the money has been vired from engineering to support science, mathematics or humanities where perhaps the HEFCE allocation is below average.
To make a fair comparison of funding, it is necessary to analyse the allocations over all subject categories in all universities. The gains and losses over all subjects in each university are added to see whether overall the university does better or worse than the average level of funding.
For full-time taught courses the list from the most generously to the least generously funded university is given in Table 1. Subject mix has no effect on this list. By taking the average funding for the subject and looking at how much better or worse a university does against the average there will be no advantage or disadvantage from having either particularly expensive or particularly cheap subjects.
Several features of Table 1 are particularly striking. First, newer universities that rapidly increased their student numbers in recent years, such as Bournemouth, Humberside, Anglia, Derby and Luton have particularly low allocations. This is a direct consequence of "fees-only" students being incorporated into the funding when the old PCFC system was replaced by the HEFCE method. Second, the new universities do slightly worse than the old universities.
On average they are funded Pounds 85 per student less generously. However, this residual effect of the binary line disappears almost entirely if the five rapid growing universities at the bottom of the list are ignored.
The most surprising feature is the huge spread in funding. There is a range of nearly 30 per cent around the average funding for teaching in 1993/94. At the extremes, Imperial College received Pounds 998 more per student than Luton University even after allowing for subject mix. When the average funding per full-time student over all subjects is Pounds 1,600, this is a large variation.
The HEFCE attempts to narrow the range of funding by applying a slightly harsher efficiency gain to the more generously funded universities. Typically the differentiation with the efficiency gain is between 1 per cent and 3 per cent. As this applies at the level of funding subjects rather than universities, the effect will be dampened and difficult to model. Crude arithmetic suggests that at this rate of progress it will take the funding council between 15 and 20 years to narrow the current range of funding to within 5 per cent of the average. Clearly this is one of HEFCE's less dynamic policies.
The results of this analysis are surprising, particularly as it seems that the new universities are, with a few exceptions, being funded as well as their more traditional counterparts. The analysis leads inescapably to that conclusion but it certainly does not feel like it on the ground.
One possibility is that if the funding of part-time teaching is less generous than for full-time then the new universities may subsidise their part-time teaching from their full-time funding. This can be tested by repeating the exercise over all subjects but this time for the funding of part-time teaching. The results, in terms of how much extra money per part time student is received by each university, can be seen on Table 2. If the new universities are looking for salvation in the analysis of the funding of part-time students then, on these results, their hopes are dashed. Indeed overall, the new universities are significantly better funded for part-time students. On average they receive Pounds 196 more per part-time student.
The range of part-time funding is even wider than for full-time with Birkbeck College receiving Pounds 1,080 more per part-time student than the University of Bath. There is no correlation at all between the funding of full-time and part-time students. Success in full-time funding is not a prerequisite for success in part-time. The University of Aston manages to move from third from the top on funding for full-time to fourth from the bottom for its part-time students.
There is one final myth to explore. It is often alleged that generous research funding is used to subsidise teaching. If this were true then a university that had done well out of the research assessment exercise might be able to operate with lower funding from HEFCE for its teaching because of a subsidy from research funds. This idea can be tested by looking for a correlation between the proportion of the total income to a university that is derived from research income against success in gaining funding for teaching. For the new universities, there is absolutely no correlation between research income and funding for teaching.
Those traditional universities that did well in the research assessment exercise and receive a significant proportion of their income for research tend also to have been more successful in gaining funding for teaching.
The new universities are not at a disadvantage, apart from the five most recent institutions where previous rapid growth is still depressing funding per student. The most surprising result is the variation in funding.
A variation of Pounds 1,000 per student between the best and worst funded university cannot be sustainable in the medium term, particularly when it is not underpinned by any policy imperative. An unkind thought is that such a variation only exists because HEFCE has not noticed it. Perhaps they will now.
Peter Knight is vice chancellor of the University of Central England, Birmingham.