The change announced on 11 April by Research England as an update to the research impact rules for the research excellence framework 2021 was introduced as “Following feedback from the sector, the initial threshold for impact case studies will be increased to 20 FTE (from 15 FTE)”, indicating that two case studies will now be required for up to 20 full-time equivalent staff submitted to a unit of assessment. This is most welcome. The original decision on staff FTEs per case study required was made late last year (“REF 2021: what are the rules and how well will they work?”, News, 26 November 2017, and Letters, 7 December 2017).
The number of impact case studies required per FTE number of staff to be returned for smaller submissions below 105 FTE in comparison with much larger submissions did not appear to be fair; moving the initial staff threshold to 20 FTE partially addresses this inequality but leaves the position that larger submissions now above 110 FTE will still be required to produce only an additional case study per 50 FTE staff returned. It is therefore surprising in making this update that the funding councils did not fully address the issue to bring the impact case study requirements back in line with those that were in place for REF 2014, which were equitable and did not preferentially favour submissions in relation to their size.
To provide a full remedy to this situation, the initial threshold would need to be moved to 25 FTE staff on the basis that the estimate determined by the REF team is that 60 per cent more staff are to be returned in 2021 than were submitted in 2014. In order to balance the reduction in the number of case studies anticipated to keep the overall numbers about the same as in 2014, which is a target to enable the impact assessment to be managed satisfactorily, the abrupt change in the quantum of submitted staff FTEs from 15 FTE to 50 FTE at a staff return of 110 FTE should then be removed and the linear relationship between staff FTE and impact case study numbers re-established.
Such an approach would also produce a more even distribution of impact case studies between the main REF panels as submissions to panels A and B are often larger than those to panels C and D, whose units of assessment, however, are likely to deal with a larger number of smaller submissions.
Pro vice-chancellor (research and enterprise)
University of Hertfordshire