REF 2029: volume measures to avoid ‘census cliff edge’ confirmed

Additional guidance has been published on the Research Excellence Framework’s new approach to calculating how many outputs and impact studies must be submitted by universities.

For the first time in the exercise’s history, universities participating in the 2029 evaluation will no longer submit lists of staff whose research will be entered for each unit of assessment (UoA), with “volume” instead determined by the number of research staff employed over a two-year period, as reported to the sector’s main data body, the Higher Education Statistics Agency (Hesa).

For every full-time equivalent researcher, institutions will need to submit an average of 2.5 outputs for a unit of assessment, though not every researcher will need to submit outputs. Previously, all eligible staff had to submit at least one output published in the REF period, though exceptions in certain circumstances could be made. The volume measure also determines how many impact statements are required for each UoA.

The change follows concerns that the previous approach incentivised universities to recruit star researchers just before the “REF census day” rather than providing an accurate picture of institutions' existing strengths.

Confirming how the new system, which aims to “avoid a REF census data cliff edge”, will work, new guidelines published on 16 January explain how Hesa will collect data based on a member of staff’s contract when calculating the volume measure.

Academic staff on either teaching and research contracts, or research-only contracts, who are on the payroll in 2025-26 and 2026-27, will be submitted to Hesa for REF inclusion, the guidance explains.

Further detail on what constitutes “significant responsibility for research” explains how these contracts must make available “explicit time and resources” for research – either in workload models or time allocations – that researchers must “engage actively in independent research” and that research “is an expectation of the job role”.

Teaching-only staff must not be submitted to Hesa, while being named as an author on a research output is not sufficient to be entered.

In some “exceptional circumstances”, research assistants who are “primarily employed to support another individual’s research rather than pursuing independent research” can be submitted if they are on a research-only contract, says guidance, which also includes a new code of practice.

Research excellence: what is it and how can universities achieve it?

Institutions have previously been asked to collect data for 2024-25, which will be used as a pilot year for the REF, though this data will not be used for REF 2029. Data for 2027-28 will not be used for the REF as it will not be ready in time for the REF submission deadline at the end of 2028, with results due to be published at the end of 2029.

Rebecca Fairbairn, the REF’s director, said the new volume measures would support “efforts to break the link between outputs and individuals”.

“The move to using Hesa data for the REF will help long-term data collection that allows the sector to explore the changing shape of UK research capacity,” she said.

“We are committed to ensuring REF 2029 is inclusive of all research-related staff and encourages engagement with practitioners and those with non-academic expertise.”

jack.grove@timeshighereducation.com