APPENDIX 31
Memorandum submitted by the UK Life Sciences
Committee
1. Do the results represent a genuine improvement
in research performance?
UKLSC believes that the results do represent
a genuine improvement in the quality of the research submissions,
and this is the view of two of our committee members who actually
served on different RAE panels and saw the evidence and experienced
the assessment process at first hand.
It is clear that there is a large amount of
excellent research of national and international standard being
carried out in UK universities, which must be funded adequately.
Universities have utilised the funding secured
in the previous RAE to improve the quality of their research,
and also used their earlier experiences of the RAE process to
maximise their chances of success. The improvement in grades has
arisen in part from institutions focusing resources strategically
on areas of research in which they can do well, and from improving
the way that they manage academic staff. It is now considered
reasonable to identify fully research active staff and to distinguish
them from staff who have a large teaching commitment, and to manage
their activities in different ways. This policy of developing
individuals' strengths would be expected to promote high quality
research and to reduce the full-time-equivalent research staff
submitted to the RAE. However, for strategic reasons some departments
may have taken this further by being selective in the choice of
"research-active staff" included in submissions. Not
all UKLSC members are content with the way that the RAE is driving
the evolution of two categories of university staff"research-active"
and "teaching-active", which they consider to be detrimental
in the long term. They argue that researchers who do not teach
may become poor communicators, and teachers who do no research
may fall behind as subjects progress and lack the experience to
train new students for a research career. For the policy of managing
academic staff according to their individual skills to be successful
it will require improved university career structures to recognise
and reward excellence in teaching as well as research.
There is some concern about the high level of
fragmentation of work in the Life Sciences between panels, which
the cross-referral process (which might not be used equally by
all panels) could not fully address. This may have led to differences
in methodology and even in standards being applied to comparable
research. The review by HEFCE of the operation of RAE 2001 will
need to assess how effective the umbrella panel concept was in
helping to avoid such anomalies.
An unfortunate knock-on effect of strategies
intended to maximise research performance for the RAE is that
activities not considered to be important for the RAE are discouraged.
It has, for example, become more difficult to recruit people to
serve on the committees of learned societies, take on editorships,
or to write reviews. This may be considered not to be in the broader
interests of science.
UKLSC would support the arguments presented
by Sir Howard Newby, Chief Executive of HEFCE, for believing that
the improvement in grades is genuine (Research Fortnight, 19 December
2001, p 17). In particular:
The RAE process was transparent and
open, with umbrella panels intended to ensure consistency in grade
allocation between related panels;
International experts were employed
to validate the highest grades awarded;
Citation analyses are consistent
with improved international recognition of British science in
the period between the two most recent RAE exercises.
2. What is the best way forward?
In a submission to HEFCE's review of research
policy and funding made in late 1999 UKLSC reported that views
were fairly evenly divided between continuing with an RAE based
on the present system, or moving to some simpler formula that
did not require an extensive evaluation exercise. The RAE was
accepted reluctantly as the fairest and most rigorous system available
(although some would question whether it would withstand a thorough
cost-benefit analysis), and for this reason many people thought
it justified the cost in time and money. One problem arising from
the government refusing to provide additional money to HEFCE for
2002-03 to fund the improved grades in RAE 2001 is that people
will think that the effort is not worthwhile. As was pointed out
by the Science Policy Research Unit (Sussex) the whole ethos of
the RAEthat if a department improves its research it will
receive more fundingis under threat.
HEFCE and other organisations with clout must
present a strong argument to the government's Spending Review
for additional money to be made available to HEFCE to enable it
to fund fully departments for their performance in RAE 2001 for
the remaining four years of the assessment cycle. The case can
clearly be made from a position of strength and success. For 2002-3
HEFCE's intention to maintain the unit of resource for 5* departments,
while providing a safety net to tide over departments graded 3b
and higher is probably the best compromise. The algorithm adopted
for 2002-03 should not be continued in future years since the
consensus view of the science community is that the present degree
of funding selectivity is about right, whereas the 2002-03 arrangement
will increase selectivity.
The present provision of research infrastructure
support is inadequate. Schemes such as the Joint Infrastructure
Fund helped to address a backlog of infrastructure defects, but
have been insufficient to meet demand. The overheads provided
by research council and EU research grants do not cover the real
cost of research, and in the life sciences the failure of the
Wellcome Trust to provide an economic overhead exacerbates the
problem. For the longer term the Higher Education sector may need
to start thinking more seriously how much research infrastructure
government funding can maintain in different universities. Sir
Martin Harris (Vice-Chancellor, Manchester University) has spoken
recently of neighbouring universities needing to come together
in a "federal university" with shared facilities and
infrastructure where appropriate in order to avoid duplication
(eg Times Higher Education Supplement, 4 January 2002).
3. What are alternative strategies for allocating
higher education research funding?
UKLSC reinforces the view expressed in the earlier
HEFCE submission that the dual support system should be maintained.
There will continue to be a need for research performance to be
assessed. The five-year cycle is seen by most people to be about
right: a shorter cycle would give insufficient time to make changes,
a longer cycle would risk relaxation of performance. However,
some UKLSC members argue that now universities have been through
several RAE cycles they have focused their research strategies
and management and the cycle could usefully be extended to seven
years. This would encourage more speculative research and risk
taking. On balance UKLSC would prefer a simpler system of assessment
than the current form of the RAE, but there is no real consensus
among members on what this should be.
The two methods mentioned most frequently are
basing the funding allocation for research infrastructure support
on the total research income from all sources, and basing it on
the quality of published work, for example using the impact factors
of selected papers:
The former method would benefit larger
departments with leading researchers bringing in major grants;
ie it would reward success and help to support truly internationally
competitive research. Since grant applications are subject to
intensive peer review it would also satisfy the requirement that
funding should go to groups performing research recognised to
be good. On the other hand infrastructure funding could be seen
to follow research that is in vogue for which it may be easier
to command grants, and more expensive research. Work in novel,
speculative areas would be penalised. Even within biology the
different level of resources required for work in different disciplines
seriously hampers the usefulness of a simple income-based formula.
An important argument for funding
on the basis of the quality of published work is that it recognises
directly the quality of research output. The difficulty, however,
is that simple publication parameters are notoriously unreliable
indicators of quality. In the biology arena, for instance, work
of comparable quality at the chemical/physical end of biology
and in a vogue area of cell biology published in journals of similar
international repute would gain markedly different citation scores
and impact factors. Impact factors also differ markedly between
the different disciplines encompassed in higher education research.
Finally, in the Life Sciences, high impact journals tend to be
American. There is already concern that the emphasis in the current
RAE on impact factors causes problems for the journals published
by the UK learned societies, internationally respected though
they are.
January 2002
|