Universities may be reluctant to enter the?subject-level assessment planned as part of England's teaching excellence framework (TEF) if they have already gained a good institutional rating, a sector expert has warned.

With the government set to announce overall TEF ratings for almost 300 higher education providers at the end of May, attention is now turning to proposals on how to assess the strengths of universities' different departments.

A two-year pilot on subject-level assessment is due to start in 2018-19, with all institutions set to take part in the more granular ratings two years after that.

However, Geoff Stoakes, head of research and special projects at the ³ÉÈËVRÊÓƵ Academy, who has been involved in discussions around the creation of discipline-level ratings,?said the government had a ¡°conundrum it needs to resolve¡± on participation in the TEF¡¯s later stages.

³ÉÈËVRÊÓƵ

Speaking at a joint?HEA-Universities UK conference on teaching innovation, he suggested that many institutions would prefer to rely on the planned three years of inflationary fee uplifts allowed by a successful institutional TEF rating for as long as possible.

¡°Why would you enter discipline-level assessment if you have already had a successful institutional submission with three years of inflationary fee uplifts?,¡± asked Dr Stoakes. He added: ¡°Why would you put that [institutional rating] at risk?¡±

³ÉÈËVRÊÓƵ

This scenario?assumes that ministers are able to reinstate the link between the TEF and tuition fee rises, which was blocked by a House of Lords amendment to the ³ÉÈËVRÊÓƵ and Research Bill.

Dr Stoakes said that the move to subject-level assessment would introduce several challenging problems for policymakers, including how to weight provider and discipline submissions.?Some TEF criteria, such as teaching and learning assessments as judged by the National Student Survey, ¡°rest more comfortably at a discipline level¡±, he said, while others, such as non-continuation rates or graduate employment outcomes, were less appropriate for use at a smaller scale.

Those drawing up plans for the TEF Year 3 pilot also had yet to decide how subjects would be carved up for assessment, said Dr Stoakes. Options include 21 subject areas or 108 more specific classifications used by the ³ÉÈËVRÊÓƵ Statistics Agency, the 36?subject areas used in the 2014 research excellence framework, or a?potential ¡°hybrid¡± involving 40 to 60 subjects under six subject groupings.

However, the desire that the TEF is not ¡°big, bossy and bureaucratic¡±, as Jo Johnson when launching the framework in July 2015, would be difficult to achieve, Dr Stoakes said.

³ÉÈËVRÊÓƵ

¡°When you move from an institutional to a discipline level assessment, this danger is compounded,¡± he said.

jack.grove@timeshighereducation.com