Do universities really face funding cuts if they fail to promote diversity?

Of course not. This morning’s coverage in The Times (and a suspiciously similar story in the Telegraph) concerns a small-scale pilot of the Research Excellence Framework (REF) 2029 component covering “people, culture, and environment” (PCE).

REF is linked to university funding, in that results (along with staff numbers) are used to determine the allocation of non-ringfenced funding (QR) for research to providers in England, and similar schemes elsewhere in the UK.

The contribution of REF to this allocation is frequently misunderstood. Broadly speaking, if research in your subject area in your institution was judged to be “internationally excellent” (3*) in the last REF you get an allocation of funding, if it was found to be “world leading” (4*) you get four times the allocation, and otherwise – if the research was merely “recognised internationally” (2*), “recognised nationally” (1*), or if it was unclassified – you get nothing.

There’s a further weighting by subject area (lab and clinical subjects get more), but provided you are within the hallowed 3*/4* circle the primary determinant of how much funding you get is really the volume measure: the amount of staff time devoted to research. To be clear, this allocation is not ringfenced – it goes to your university to spend on what it wants, not to your department or subject area.

England allocates about £2bn each year to QR funding – you may wish to compare this to £44bn of sector income (from all sources, public and private) each year, or to the £21bn higher education providers spend on employing staff. QR is a useful and worthwhile allocation but it is a drop in the ocean – I always love telling people that the majority of universities earn more from their residential and catering operations than their research (indeed, the total income to the sector from basically being a hotel was higher than the total income from QR).

Your REF grade – that tiny component informing the allocation of this small by sector standards income stream was made up in 2021 of three components:

• Research outputs (the quality of the work you publish or otherwise put out into the world) – 60 per cent of your score
• Research impact (whether your research does any good to the wider world) – 25 per cent of your score
• Research environment (the people, equipment, support, and policies you have to support researchers) – 15 per cent of your score

As you can see, the biggest chunk of your overall score is the “traditional” research assessment: your academic peers from around the world engage with the research outputs you submit, and put them into one of the categories described above. There’s an entire literature as to whether this is a good use of time and money, but we are where we are.

Another thing worth noting here is that, for QR allocations, there is actually an allocation for each of these three components (using the same weightings) that gets mashed together at the end. So it is theoretically possible to get an allocation for your subject area based only on the quality of your research environment

For the 2029 REF, the current proposal is to weight three components as follows:

• Contribution to knowledge and understanding (quality of your research outputs) – 50 per cent
• Engagement and impact (how the outside world engaged with your research) – 25 per cent
• People, culture, and environment (the people, equipment, support, and policies you have to support researchers) – 25 per cent

The proportions refer to the contribution to your overall REF score – there has yet been no indication that Research England would follow these proportions in allocating QR (and any change along these lines would apply for 2030-31 allocations at the very earliest).

This has been a long way round, but this morning’s coverage is just the latest salvo in a small number of people being upset that the proportion of the total mark attributable to the research environment has risen by 10 per cent, while the proportion of the total mark attributable to the quality of your research outputs has fallen by 10 per cent.

If you consider the purpose of REF and QR, you would have to conclude that it is a mechanism to ensure that our universities remain good at doing research. Although income from QR is used in different ways at different providers, the general sense is that it represents an investment in research capacity – and thus, the research environment (the resources and support you need to do good research). Neither your REF score or QR allocation are prizes for writing the “best” research article.

Whoever has briefed The Times is upset about aspects of the recently published PCE pilot exercise guidance. The REF team is still very early (arguably too early!) in the process of designing how PCE assessment will work – the current exercise gets willing volunteers to make a mock submission of PCE-related “stuff” to help the team understand what evidence for a good research environment providers can muster and if any of it is any help. It’s not a full submission either – the already existing stuff (covering research income, research facilities, and research infrastructure) is not part of the pilot because we already know how to do that. It will be a big part of the final submissions to REF2029’s PCE components.

The REF team asks for evidence across “five factors that enable positive research culture”. These are:

• Strategy – do you actually have a plan to support researchers and research?
• Responsibility – is your research ethical? Does it have academic integrity?
• Connectivity – do you work well with other teams, providers, and subject areas? Can people actually access and draw on your research?
• Inclusivity – is your research environment collegial and inclusive? Are there any barriers (other than merit) to being able to contribute?
• Development – do you train and develop the people engaged in performing and supporting research?

There’s no complaints about most of that – it is generally expected that these things (Ethics! collaboration! looking after your people!) are pretty much essential if you want a high-performing research team. I mean, you could lock all of your researchers in a cupboard and yell at them till they publish good papers, but I feel like you wouldn’t see much success.

Of course, the culture warriors are performatively upset about the “inclusivity” component – which at this stage must be worth substantially less than 5 per cent of your total REF score (remember the existing environment measures from last time round aren’t even in this pilot).

Each of these components has a number of possible indicators linked to it in the submission guidance. Some of these may end up being mandatory, most will not – we don’t know yet, that’s why we’re running a pilot. Here’s some examples of suggested indicators for the “inclusivity bit”:

• Data about researcher training – is it any good?
• Do you give people proper jobs or put them on fixed term contracts? Do you pay them properly? Do you assess their performance fairly and transparently?
• Do you have a harassment policy? Do you have a whistleblowing policy? How do you deal with bullying and bad behavior?
• What does your department look like – is it all middle-aged white guys or are you more into having a diverse set of perspectives and experiences?
• There’s loads more options, I won’t list them all.

I read the list several times – there’s nothing that makes you declare a political allegiance to the “woke mob” that I’ve spotted. Maybe if you get upset about the idea that people other than middle-aged white men can do good research – and maybe if you think that what must be by this point a fraction of one percent of your overall REF score would be affected to the extent that you would lose the tiny amount (outside of a handful of massive universities) that most people get from QR… in that case, maybe just maybe the article in The Times has a point.