The Committee for Academic Freedom (CAF) is preparing a formal response to proposals that could see universities rewarded for embedding contested approaches to equality, diversity and inclusion (EDI) in their research strategies, amid concerns about the consequences for academic freedom.
Research England, which manages the Research Excellence Framework (REF) on behalf of the UK’s four higher-education funding bodies, is seeking final feedback on draft assessment criteria before they are finalised later this year. The REF is used to assess research across UK universities and helps inform the distribution of around £2 billion a year in public research funding. Because REF results help determine the distribution of substantial public research funding, they also carry considerable reputational weight as a benchmark of universities’ research performance.
Under REF 2029, research outputs will account for 55% of the overall assessment, engagement and impact for 25%, and a new, expanded Strategy, People and Research Environment (SPRE) component for the remaining 20%.
SPRE assesses the institutional and departmental conditions in which research is produced, including universities’ strategies, research culture, support for staff, infrastructure and collaboration. Research England is currently seeking feedback on its latest draft assessment criteria.
It is this part of the proposals that concerns us.
The latest draft SPRE criteria say that “each section” of universities’ submissions should demonstrate their approach to EDI. Assessors will also look for integration between institutional and unit-level strategy, including evidence that institutional strategy has shaped the “activity and direction” of individual academic units.
There is an obvious risk here, given that universities have powerful financial and reputational incentives to maximise their REF scores. If demonstrating particular approaches to EDI becomes part of what REF rewards as research excellence, universities may in turn acquire an incentive to make decisions about research priorities, funding and support according to what they believe REF assessors expect to see.
The trouble is that EDI can mean rather different things in practice.
At its most legally grounded, EDI can mean little more than compliance with equality law: fair treatment, non-discrimination and equality of opportunity. But universities increasingly also adopt much thicker conceptions of EDI, drawing on contested ideas derived from critical race theory, decolonisation, gender identity theory and other related bodies of thought. CAF takes no position on the merits of those frameworks, except to point out that, like any other social or political theory, they are not requirements of equality law. As such, they remain perfectly legitimate subjects of academic inquiry and disagreement, and academics must remain free to question them in, and as part of, work that may itself ultimately be submitted to the REF, rather than being expected to subscribe to them.
That this distinction has not always been observed across the sector is apparent from the way universities have frequently operationalised broader conceptions of EDI, not only through requirements to decolonise curricula or definitions of “transphobia” extending to contested but lawful beliefs, but also through external benchmarking and award schemes such as Stonewall’s Workplace Equality Index, Athena Swan and the Race Equality Charter. In some cases, external EDI guidance has resulted in institutional policies or expectations embodying not only contested propositions, but propositions founded on misstatements of the law.
Unfortunately, we cannot simply assume that REF means the thinner conception, since its own guidance says that the Codes of Practice which all institutions wishing to participate in REF 2029 must produce “go beyond procedural and legislative compliance”, while section 8 of that guidance states: “excellence in SPRE is expected to extend beyond the minimum standard”.
To be sure, REF 2029 says that SPRE “fully respects institutions’ existing legal duties and sector commitments relating to academic freedom and freedom of speech” and that research environments should support “a diversity of scholarly perspectives”. But legal duties exist within a complex, interlocking statutory ecosystem. Acknowledging the lawful boundaries of academic freedom is therefore of limited value if another part of the assessment framework is premised on a conception of equality broad enough to incentivise conduct that intrudes upon those boundaries.
At present, the draft criteria do not make sufficiently clear which substantive conception of EDI assessors are meant to reward, by what standard they are supposed to assess it, or whether they possess the expertise to distinguish the requirements of equality law from contested EDI policies and theories. That is distinctly problematic because, without any corresponding instruction to assess EDI in a way that protects lawful academic disagreement, an expansive conception of EDI risks allowing the assessment of equality-related matters to overreach and impinge upon English universities’ statutory duty under the Higher Education and Research Act 2017 to secure freedom of speech within the law, including academics’ freedom to question received wisdom and advance new, controversial or unpopular ideas and opinions.
Those are precisely the freedoms that risk being put under pressure if contested propositions associated with thicker conceptions of EDI – for example, the primacy of gender identity over biological sex in areas of public policy, or mandatory approaches to decolonisation – are themselves treated as features of an “excellent” research environment, while the protection of academics whose scholarly output challenges them is not.
None of this means that a REF panel will directly punish an individual academic or mark down a controversial article, not least because REF 2029 deliberately decouples individual academics from submitted outputs. The potential threat to academic freedom outlined above is subtler, more insidious and less visible.
But decoupling individual academics from submitted outputs does not decouple them from the institutions in which their research is produced. Universities still control research time, internal funding, workloads, promotions, appointments, institutional support and strategic investment. If those institutions have an incentive to adopt whichever conception of EDI they think will score best – including the thicker and more contested conceptions already found in university EDI policies and practices – individual academics inevitably work within the resulting research environment. That could leave those whose research, conclusions or beliefs sit awkwardly with institutional EDI orthodoxies working in materially less supportive conditions, even though their scholarship is entirely lawful and academically meritorious.
REF has always been controversial among academics, not least because of the difficulty of reducing something as complex and intangible as research excellence to an institutional assessment exercise. But if the system has any justification, it must surely be that it identifies and rewards excellent research, and the conditions that genuinely enable it. Equality rules should ensure that researchers are treated fairly, not create incentives for universities to make conformity with contested EDI objectives a condition of research support.
CAF will be responding to the current REF 2029 survey and asking for an explicit academic-freedom safeguard before the criteria are finalised. We think the problem could be substantially reduced by a relatively modest amendment. The most problematic section of the draft guidance currently states:
The institution’s/unit’s approach to equality, diversity and inclusion: statements should include approaches to promoting equality through the active inclusion of a diverse range of staff and research student roles and voices, particularly for under-represented groups, in the development and implementation of their strategy.
We propose instead:
The institution’s/unit’s approach to equality, diversity and inclusion: for the purposes of assessment, credit should be awarded only for measures consistent with applicable equality law and with duties relating to freedom of speech and academic freedom, including measures promoting equality of opportunity, non-discrimination and the active inclusion of a diverse range of staff and research student roles and voices in the development and implementation of strategy. No greater or lesser credit should be awarded by reason of an institution’s adoption of, adherence to, or rejection of any particular political, philosophical or ideological framework. The active protection of lawful academic disagreement with such frameworks should itself be recognised as evidence of an excellent research environment.
Research England’s survey allows organisations just 1,500 characters – barely 250 words – for additional comments on the entire proposed SPRE approach, despite the complex and potentially far-reaching issues it raises.
We do not think that is remotely adequate for meaningful scrutiny of proposals that could influence research priorities and institutional behaviour across the UK university sector. Our proposed amendment alone, for instance, runs to roughly 118 words.
We will therefore also be writing separately to Research England setting out our concerns in full. If the final criteria fail adequately to protect lawful academic disagreement and academic freedom, CAF will consider seeking legal advice on whether the final framework is amenable to challenge by way of judicial review.