The UK has an outstanding higher education system delivering excellent teaching so universities should approach the proposed Teaching Excellence Framework with confidence.

This excellence has been demonstrated over many years, through one of the longest running external quality assessment systems in the world. The system, now being reviewed by the Higher Education Funding Council for England in parallel with a government consultation, should be the starting point of any TEF. It makes perfect sense to bring the TEF and a reformed quality assessment system together — the system providing the bedrock, the TEF the nuance.

While a culture of developing outstanding teaching is both supported and facilitated at institutional level, it is delivered by departments, so any TEF will need to consider both. At institutional level it should consider wider strategies, policies and practices as well as processes for reward and recognition.

However, real assurance about teaching quality will need to happen at the level of department. Even in the best institutions, the quality of teaching across departments varies. Differences within an institution may reflect not only variations in quality but also different approaches by different disciplines, and the backgrounds of the students. Jo Johnson, the universities minister, recognised this recently when he emphasised the need for any TEF model to integrate widening participation.

These differences within institutions mean that data used for assessing quality will need to be benchmarked against a variety of contextual information, and institutions will need to be able to provide an accompanying narrative, reviewed by a panel of peers including students and employers.

After months of discussion about how you actually measure teaching excellence, and whether it is to do with teachers’ performance or distance travelled by the learner (so-called “learning gain”), a consensus is growing that a range of proxies pointing to high-quality learning experience, student outcomes and excellent teaching will be needed. But it has also become clear that some of these proxies do not currently exist. Nor should the TEF be over-complicated, bureaucratic or burdensome.

It will therefore need to evolve as more appropriate metrics and proxies become available. We don’t want to undermine the TEF by using metrics that are not seen as being true measures of excellence, and we don’t want to undermine the reputation of UK higher education by setting a high pass-rate for the TEF, with only a small proportion of institutions passing. There is a strong case for any system to be based on institutions passing – or not – according to whether or not they meet certain criteria, rather than how they measure up against each other.

We won’t be able to get it 100 per cent right in year one; the important thing is not to get it wrong. We need to take as our model the way the Research Assessment Exercise and its successor, the Research Excellence Framework, have continued to evolve over more than 20 years.

This is particularly the case as the timeline proposed by the Department for Business and Skills is very ambitious. It includes differentiating between institutions and allowing some to raise fees by inflation and others not, depending on their TEF performance, which will demand robust and agreed criteria or risk lengthy appeals. These criteria may take time to develop. There is a strong argument to be made for the simplest possible process in year one before developing the TEF further in year two and beyond.

This article first appeared in HE from Research Fortnight.

Alex Bols is deputy chief executive of GuildHE