James Hand, University of Portsmouth
The UK’s higher education funding bodies announced initial decisions for the operation of REF2028 on 15 June 2023, thirteen months after the REF2021 results were announced. In some respects, we are already over 40% through the likely new cycle, with REF2028 submission due in 2027 (and guidance and criteria due be drafted and finalised in 2024). This short post looks at the emerging rules (some of which are open to consultation until this Autumn, some will be subject to further elaboration later and others are in need of clarification) amid the context of the REF2021 Law submissions.
The Unit of Assessment structure is likely to be the same (subject to representations about disciplinary developments), with Law currently one of 34 subject-based Units of Assessment. REF2021 saw 69 submissions to the Law unit sub-panel (with nine new submitting units and seven which submitted to REF2014 not submitting to Law that time (Panel C Overview, p.104). Articles in law and criminology journals were also submitted to the likes of UoA 17 Business & Management, UoA 19 Politics and International Studies, UoA 20 Social Work and Social Policy, UoA 21 Sociology, and UoA 25 Area Studies.
Other similarities include the primary use of expert review by the sub-panels, an output multiplier of 2.5, the likely presence of a Covid impact statement and a focus on interdisciplinarity (but not using the interdisciplinary flag on outputs as its use by HEIs had been too inconsistent). Furthermore, the Codes of Practice which each HEI drew up for REF2021 are intended to be refreshed on a light-touch basis for REF2028 (though the initial decisions document recognises that some institutions ‘may wish to make more substantial changes… building on learning from REF2021’, para 84) and so the (somewhat varied) approaches to significant responsibility for research look likely to remain. The assessment will also continue to be based around three elements – formerly Outputs, Impact and Environment – but these will be renamed, reweighted, and refocussed on a broader range of activities.
The changes are intended to place greater weight on the research environment while still recognising that outputs, in particular a diversity of outputs, are of central importance. Outputs will fall within the assessment of Contribution to Knowledge and Understanding which will comprise 50% of the overall assessment (down from 60% for outputs). Impact will remain at 25% and be joined by Engagement, with People, Culture and Environment comprising the other 25%.
Contribution to Knowledge and Understanding
At least 10% of the 50% (which reading across the other headings means at least 5% of the overall assessment) will be for a structured explanatory statement which: will present the wider contribution to knowledge in the area; may showcase the variety of research undertaken (including that not submitted, and that which supports others such as making data available); and shall show how the submission is representative of the research and researchers in that area in that HEI. The precise weighting and content will be subject to consultation with the main and sub-panels and there will be consultation with the sector on appropriate metrics/indicators.
There will no longer be a census date on which staff have to be employed; instead the volume of outputs will be determined using an average FTE of eligible staff during 2025/26 and 2026/27 for REF2028 (and the whole REF cycle thereafter) using HESA data. It is intended that the link between outputs and individuals (partially delinked in REF2021) will be removed: there will be no minimum requirement per individual and no maximum (subject to the demonstration in the explanatory statement that the submission is representative of the research and researchers); there will be no individual circumstances to take account of (which may give rise to the concerns outlined ahead of REF2021 here and here) but unit circumstances may still be recognised; and outputs can be by any staff member with a ‘demonstrable link’ or ‘substantive connection’ (including teaching fellows or non-academic staff). There is no mention of staff being able to bring outputs with them to an institution (which was lately allowed on a transitional basis for REF2021 due to the effect on, particularly, Early Career Researchers).
It is intended that the rules will encourage a wider range of submission types and reflect the wider research environment. Compared to the Panel C average, Law already has a more diverse range of outputs with four-times the proportion of book chapters submitted (at 16% of outputs) and twice the number of books (at 23%) and consequently fewer journal articles, although at a still dominant 58%. Articles from over 900 journals were submitted to the Law sub-panel with 20% of submission articles coming from the top 10 most submitted journals with just over five-sixths of the outputs being in journals submitted five times or fewer. There is thus some diversity of output, albeit skewed, and looking more broadly, outputs such as conference contributions, research reports, and working papers generally (with one or two exceptions) made up fewer than 1% each of outputs across Panel C (and Law is no exception there). These output types (and more specialist journals) may come more to the fore under the proposals either through the output submission itself or the statement.
Engagement and Impact
This element will retain the case studies but, as with outputs, there will also be an explanatory statement (similar to the REF2014 impact narrative / impact part of the REF2021 Environment statement). This narrative is proposed to have a variable weight depending on the number of case studies, ranging from 20% of the sub-profile to 50% for the smallest submissions. The requirement for impact to be underpinned by research of 2* quality will be scrapped, to encourage a broader range of impact to be recognised, but there would still need to be underpinning research (i.e. a contribution to knowledge but one which does not need to be so rigorous or significant). A third criterion, of rigour, may however be added to reach and significance when assessing impact so as to provide focus on the process as well as the outcome.
It is proposed that the number of case studies required will be reduced in some cases. For UoAs with fewer than 10 staff, only 1 case study would be required. For those with over 160 the new figure would be 7 rather than 10 under the old rules. However, the effect is uneven. For UoAs with 10-19.99, 20-34.99, 40-49.99 and 60-64.99 FTEs there is no reduced burden in terms of numbers. The reduction only affects UoAs with staff numbering under 10, between 35-39.99, in the 50s and 65+. If this had applied last time, all else being equal, nearly two-thirds of Law UoAs would have had no reduction.
People, Culture and Environment
The content here will move from a narrative statement to a more defined, questionnaire-style template and specific content and metrics are to be consulted on later. It will comprise a disciplinary-level statement and an institutional-level statement with the institutional-level statement making up at least 20% of the sub-profile score (the precise amount to be discipline-specific). However, it can be seen that narrative elements of the old Environment Statements will now fall to be assessed via the explanatory statements to be introduced under the above elements.