Some first thoughts on the initial REF 2028 decisions

I've been asked by several journalists to comment on the initial decisions on the shape of the next UK Research Excellence Framework published today.

As usual, I wrote some notes up and wanted to share those first reactions. I'm not going to summarise the announcement, just reaction to it, but Jack Grove at THE has written a great short summary here. For more context you should also see this excellent reflection by Wilsdon, Curry and Gadd at Research Professional.

This feels on first reflection like a big step in the evolution of the REF away from an exercise focused on evaluating the research performance of individuals as a means of assessing UoAs - with all the perverse incentives that that created - to a system that aims to take a more balanced look at research quality, culture, and impact. The idea of modifying REF so it makes it more feasible for people to move in and out of academia is particularly welcome.

Reducing the weight of research outputs in the overall score, removing the max/min requirement per researcher and introducing a narrative/self-evaluation statement about the output profile are welcome steps towards addressing some of these perverse incentives - but the prevailing cultural expectations of productivity within universities are not only driven by REF but also chime with traditional academic prejudices, and are so baked in to the way university leaders and mentors think and act that it may take quite some time to shift behaviour

This is going to be particularly the case where something new is introduced - so the introduction of a new research outputs statement that will account for a significant proportion of the research output score - even if quite a structured template is provided - will lead to lots of scrambling for insights into what is expected, and this will be subject to folk wisdom among research deans, rumours and claims of unique insight from consultants offering help write the things for a fee.

Introducing a new statement for the impact strand of REF should allow for consideration of wider impact and engagement that doesn’t fit the existing rather linear impact case study model, and enhancing the statements for research environment and culture at the UoA and institution level is also welcome. We should see this not just part of summative evaluation but as a public statement of intent that can act as a implicit contract between the institution hosting the UoA and the public funders that the institutions can be held to by their researchers and the research community. It is beyond time that universities are required to live up to their statements about research culture and support.

A big question remains about how the REF scores will be used in the future. Although REF presents itself as a general health check on the UK research system, it exists to input a score into a funding formula for block funding that itself exists in order to drive the concentration of research. The current government’s modest commitment to rebalancing public R&D spending outside the Golden Triangle is not consistent with this institutionalised imperative to increase concentration over time but we haven’t heard anything about how future REF scores will be used instrumentally in the funding system.