Next steps for REF? We need to repair the sector’s health first

This post accompanies a talk at the Westminster Forum Projects | Next steps for the REF – independence and positive research environments, delivering and measuring impact, and the future of open access event, 23/3/21. Slides available here

A while ago I was invited to speak at the Westminster Forum in a panel session entitled “Research environments in the REF – stimulating positive cultures and wellbeing, academic independence and interdisciplinary research“. When I first accepted the invitation we were pre-COVID, some time ahead of the REF submission and the prospect of talking about ‘next steps’ seemed eminently sensible. However, with the rescheduled conference now clashing with the final throes of REF (no criticism, simply an artefact of REF date extensions + challenges of arranging a conference in the midst of a pandemic), I find my mindset has changed. Not that we shouldn’t think about next steps, but because if we don’t take stock of the damage across the sector first, we can never really reach a point of wellbeing.

Before I start, it’s important to note that we shouldn’t pretend that REF can be blamed for everything – that would be an immensely simplistic and scapegoating way to assign all the ills of the sector – but with an intentional focus here on REF and impact, it’s essential that we acknowledge the collateral damage felt by so many. REF is undoubtedly a double edged sword, certainly for impact; it drove the need for jobs in this space (my own included) and legitimised those working in more ‘applied’ fields, yet simultaneously formalised and scrutinised impact to an arguably harmful level. Impact has been, to a very large extent, conflated with REF, and whilst the broader impact pilot light hasn’t gone out, impact strategies are now immeasurably flavoured by anxieties about ‘what counts’ and ‘what’s biggest’. We talk about impact as a whole, yet screen out the weaker chaff from the stronger wheat to maximise our chances of income. Whilst that seems an enormously sensible strategy for an institution under assessment, it takes no account of the damage and disenfranchisement of those not picked for the Case Study team. Impact is for everyone. Go do impact. No not that way, that’s not enough. Move aside for those doing better stuff. As much as we’d like to pretend we don’t, we still trade off impact star players for our cases with no recognition of how many others were put on the subs bench.

REF has introduced terms into our academic lexicon we will struggle to unlearn. Outputs, impacts and people are appraised in terms of how ‘REF’able‘ they are. Evidence has – much to the chagrin of my international counterparts – become both a verb (‘can it be evidenced?’) and noun (‘we need the evidence’). Yet its language legacy is not matched by sustained capacity or expertise. A 2020 survey led by ARMA showed that 58% impact personnel were on short term contracts, with 72% contracts finishing at the end of REF. 72%. We grew an army of people to deliver REF impact, now or soon to be disbanded, with those left burned out and wondering how to re-energise a tired and distrusting sector.

I talk routinely about the need for impact literacy (the understanding of impact) and institutional health (the infrastructure needed to support healthy practices). However these these need to take a temporary backseat before thinking about ‘next steps’ whilst we recognise how the sector is feeling. I’m aware that focusing on ‘feelings’ may appear to be a superficial and transient indulgence given sectoral pressures to obtain ever more reduced funds, but if we don’t genuinely take stock and understand why such committed people are so burnt out, so despondent, we will not only lose vital knowledge and skills, but also irrevocably stain the relationships between academia and society.

The sector is not well

Ahead of the talk I reached out to colleagues and was saddened, yet not at all surprised by their level of despondency. Within impact, people who have fought so hard over the years to drive a positive impact culture, now exhausted and planning to leave their job or even the sector. Tired of the narrowing of impact to page length and font compliance. Exhausted by the discord in rhetoric between ‘impact matters’ and ‘only if it’s big’, and disillusioned by the tensions arising from conflicting rules and disparity between weightings for impact and the underlying environment. It says it all that when I asked them for images to illustrate REF, I received pictures of burning buildings and frayed rope. I also reached out more widely to colleagues in the academic community* to invite comments on ‘next steps’ for REF, and was inundated with stories of demotivation, damage and despondency. There’s no way to do justice to the extent or depth of these issues, but are perhaps best encapsulated by one comment that “the damage done perpetuates many harms and maintains toxic working practices”. Issues include:

  • Inequalities cemented and deepened; those with capacity to work longer hours, travel, physically well and with no care responsibilities are more able to meet REF-related progression criteria and thus ‘climb the ladder’. Those who can’t, including part time academics, disproportionately struggle
  • Academic methodologists and non-research staff made invisible, their work pivotal for, but omitted from accounts of impact glory.
  • Anxieties related to rule interpretation, risks of accidental non-compliance, second guessing reviewer expectations, seeking to perfect cases without knowing what ‘perfect’ looks like, and marrying authenticity of accounts within rules and template space.
  • The making of an unrelenting engine; Excessive administrative burden, substantial time demands beyond standard workload, continual internal deadlines, multiple iterations of cases and review points, excessive process time and energy, all of which prevent full consideration of the consequences of decisions taken.
  • Disciplinary disprivilege; Despite recognition of subject-based differences in the relationship between research and impact relationship, certain kinds of research/impact remain privileged by the exercise (eg ICS template unsuitable for more iterative participatory or practice based research)
  • Disillusionment; early optimism that social engagement would be valued (alongside outputs) swiftly replaced with despondency over requirements to instrumentalise research and commodify partnerships
  • Pausing rather than promoting research; Instructions to intentionally delay publication when there’s already ‘enough’ for REF and wait for the next cycle.
  • Bullying, harassment and damage to mental health, limited support (worsened by COVID). Stories of REF being used to “threaten, control, shame and otherwise exploit workers”, with people made to feel inadequate or a “failure” if their work isn’t included.
  • Contractual precarity and employment barriers; Short term contracts, teaching-only contracts, blocks on appointments or roles extended only so long as to complete a case study
  • Short termist REF framed approaches: institutional strategy scheduled in REF cycles, with research value conflated with its value within assessment
  • Overall: The efforts of trying to manage, negotiate and de-toxify these issues

Beyond the need to address these fundamental problems, colleagues also called for:

  • Practical necessities; clear and non-contradictory assessment guidance needed sooner, reduced scale of bureaucracy to learn
  • Extending focus; on team science, including those not on research contracts (techs etc)
  • Fuelling positive research culture not just assessing research environment
  • Embedding meaningful approaches to and measures of EDI
  • True recognition of interdisciplinarity
  • Support for early non-academic engagement without expectation of a specific return
  • Focus on systemic inequity, with resources focused on coaching and support
  • Recognition of the consequences of midstream funding cuts (eg. ODA projects)

I hear fairly routinely the phrase ‘keep going, nearly there’ at the moment (ie. ahead of the 31st March deadline, just over a week and counting), and have done for months. Positioning REF as some kind of endurance race with an inevitable sense of relief and doubtless a celebration event or two. Doubtless this motivational chant is meant well, and for many is an accurate homily, but this belies the deep scars and potentially undoable damage for many. Are we really upholding the principles of social good by wearing down the people who fuel its development? The academics whose knowledge underpins change. The impact specialists and research managers who sit alongside, intermediating between a drive for social change and compliance with assessment rules. Disregarding the real-world effects on colleagues tasked to make real world impact? Is there genuinely a belief that assessment doesn’t change impact behaviour? Impact cannot just be positioned as academic duty, nor having ‘no impact’ considered some sort of defiance of sector expectations. We’ve traded too long off the motivation of people want to make a difference, but the personal toll doing that whilst meeting requirements for every other academic monolith is just too high.

The need to repair

It would be of course overly idealistic, and arguably impractical to simply stop assessments, particularly as they do offer at least a scripted and largely transparent process to allocate public funds. It is similarly simplistic to blame university management when there are many examples of supportive and inclusive practice. There have always been philosophical debates about ‘what counts’ and what is ‘excellent’, particularly across disciplines, so a one size fits all approach cannot fit everyone, nor am I advocating an oversimplified alternative. There are noises that the future won’t simply be REF mark 3, but actually look to address some fundamental dilemmas about how we assess research. That is an immensely welcome prospect if true. But to what extent is there really going to be flex in a system ultimately reportable to Treasury? Reducing meaningful sector engagements into comparative and scorable scenarios, with results not upgradable for 7 years, is a continuingly troubling pressure on an already exhausted sector

The equation that gets us to a healthier position must include new variables. Thus far there has been dangerously little consideration of the resource burden on universities and the toll on people, with rhetoric idyllically expectant that universities can just ‘cream off’ the best examples of impact. However, this misses several fundamental points.

Firstly, the rule book(s) for REF runs to hundreds of pages, across multiple documents and weaved into multiple FAQs. Even where universities can ‘cream off’ the best cases, the necessary checks and balances requires people to develop an expert level, legalesque memory of specific points of guidance, where to find it, and to what extent it is mandatory (vs. open to interpretation). By way of clear illustration of complexity, Dr Anthony Atkin (University of Reading) recently mapped the multiple checkpoints needed to determine a single point on eligibility:

The Spider’s Web of REF impact rules. Dr Anthony Atkin (ARMA Protagonist Winter 2019)

Secondly, particularly for the smaller universities there is not simply a ‘pool’ of strong cases to draw from. If we need two cases, we have to create two cases and often cannibalise resources from elsewhere to do so. Rather than cast for the biggest impact fish, we have to set in motion a full engine of activity to get membership to a suitable pool. The capacity burden on institutions where research – or departments – are much more newly instated is far in excess of that needed for longstanding, socially partnered and challenge led initiatives` already underway.

Thirdly, assessment, or more specifically the curated, sanitised, and positivist cases created for submission, creates a false sense of dyadism between knowledge and application. Research does not simply catalyse into impact. There has been a tendency since 2014 to use the Impact Case Study database as an exemplar dataset, displaying countable effects on policy, society, the economy and more. But these obfuscate what doesn’t work, how much effort is wasted or otherwise screened out of the final story. The sector becomes held against an unhealthy benchmark of achievement, in much the same way that photoshopped celebrities drive an unhealthy view of ‘what’s beautiful’.

For many of us whose roles extend beyond REF, the task ahead of us is immense. Patching the wounds of this REF, disconnecting the now conditioned response between meaningful impact and evidential compliance, and doing so as our own attitudes to impact are at best diluted. A post REF future must recognise the ghosts of REF past. ‘Next steps’ cannot presume either a blank canvas or a sector somehow warmed by their achievements thus far. We need impact literacy. We need institutional health. We need to remember what impact is truly about and mentally and practically unbind it from REF.

The sector is reeling in so many ways, and there’s no way to do justice to the issues in a single post. But I do know this….

We need a break. We need to learn from the past. And we need to repair.

*with special thanks to WIASN for offering such important and candid commentary.

2 thoughts on “Next steps for REF? We need to repair the sector’s health first

  1. Really thought provoking Julie, thank you for sharing this important view point. I wonder how your reflections could inform work to improve research culture and bureaucracy in research – food for thought!


    1. I’d certainly hope it does inform it – I’m very aware there are a range of surveys (etc) on associated things, eg attitudes to REF, and we need to draw it all together. My concern is that these profound implications for people are masked by aggregated consultation and feedback.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s