The Research Excellence Framework (REF), formerly the Research Assessment Exercise, has become a deeply embedded aspect of life at UK higher education institutions (HEIs) but, throughout its 30 year history, it has remained resolutely controversial. While its positive contributions to driving up research quality have been acknowledged, the huge amounts of time and money invested in it have been described as disproportionate to the benefits. Furthermore, the REF is widely blamed for creating perverse incentives in the research system that are damaging to research excellence.
In response to these concerns, in December last year the Government announced that one of the REF’s biggest critics, Lord Stern, President of the British Academy, would lead a review of how the REF works and whether it is fit for purpose. This provides a welcome opportunity to re-examine the point of the REF and therefore how, or indeed whether, it should be conducted. The Royal Society has recently published its submission to the Stern Review’s call for evidence. This addresses this basic question– outlining the principles that should shape the research landscape and the role the REF can play in supporting these principles. It is intended to be the beginning of a discussion rather than a final proposal.
The dual support system for research
The REF is an integral part of the UK’s dual support system for research. Dual support combines funding awarded by Research Councils and charities for investigators to conduct specific research projects, with funding awarded as a block grant to HEIs that can be used flexibly in to deliver the infrastructure and environment needed to support the HEI’s research strategy. This system is widely acknowledged to be a key strength of the UK research landscape, as highlighted in the recent Nurse Review of the Research Councils. The dual support concept should therefore be at the heart of the REF process. While the Research Councils and charities award research funding based on the quality of specific research projects and the track record of the investigators proposing them, the REF should focus on funding institutions in support of their strategy to provide an appropriate research environment for the conduct of such projects.
The Royal Society view is that the core components for such an environment are:
Excellence in research, underpinned by the freedom to pursue intellectual interests
Collaboration to generate novel approaches to tackle major local, national and international challenges
Diversity in people, disciplines, institutions, sectors, locations and funders so that research benefits from a range of approaches
Openness and engagement with diverse audiences to increase transparency and support the widest possible dissemination and honest discussion of research outputs
Agility so that ground breaking ideas and proposals receive appropriate levels of funding and infrastructure, facilities and equipment are of sufficient quality and scale to support cutting edge research.
The Stern Review provides an opportunity to develop a REF system that supports a research culture that embodies these principles, thereby delivering high-quality, ethical and valuable research, minimizing counterproductive pressures and incentives.
Time for a more institutionally-focused approach
To achieve this goal, the Royal Society advocates a more institutionally focused REF system. Many of the criticisms of the current system can be attributed to its focus on individuals. Individuals are submitted (or not) and each individual must contribute four research outputs. Funding is then allocated based on the number of individuals submitted. This approach has resulted in HEI performance management systems that focus on perceived REF criteria, and investment decisions that support submission of the maximum number of individuals who meet these criteria. As a result, important aspects of research culture such as the work of technology specialists, work on large collaborative projects, and mentoring and support for post-docs and PhD students can be undervalued. While many of these aspects are assessed in the current REF, the focus on individuals and their linked research outputs results in an under-emphasis on these crucial wider aspects.
In the Society’s response we set out plans for a more institutionally-focused system where institutions, rather than individuals, submit outputs to be assessed by discipline specific panels, in the context of a wider institutional level research strategy. This would better reflect the REF’s role in determining QR funding allocation to institutions and it would help to foster a research culture in line with the Society’s principles for a strong research landscape.