Late last year, Alumni for Free Speech (AFFS) published a report that found a statistically significant correlation between certain types of equality, diversity, and inclusion (EDI) related free speech compliance failures at UK universities, and levels of university spending on EDI.
AFFS is an alumni organisation headed by senior lawyers campaigning to improve universities’ compliance with their free speech legal and regulatory duties.
In brief, the AFFS report contains information on certain UK universities’ levels of spending on EDI – obtained via freedom of information requests – alongside analyses of universities’ compliance, or rather non-compliance, with their free speech legal duties, conducted by AFFS and with reference to guidance issued by the higher education regulator, the Office for Students.
It also includes a linear regression analysis finding a statistically significant correlation between universities’ levels of EDI spending and certain forms of free speech compliance failures, together with qualifications about certain limitations of the report.
What the numbers show
The AFFS report finds that higher levels of EDI spending correlate in a statistically significant way with higher levels of certain types of EDI-related compliance failure. As the report states:
Correlation does not, of course, establish any necessary causal link between levels of EDI spending and free speech compliance failures at a university.
However, as the report explains, the study controlled for both the size of institution – measured by the number of staff – and Russell Group membership, which correlates strongly with the quality of research environment. Neither total academic staff nor Russell Group membership had a statistically significant effect on the correlation between EDI spending and compliance failures.
One recently proposed explanation of the correlation – that “institutions with higher EDI investment are overwhelmingly large, research-intensive universities with more complex staffing structures and regulatory exposure” – doesn’t appear to be congruent with the data.
So while the analysis doesn’t allow us to establish a causal relationship between EDI spending and non-compliance, it does show that the two correlate, and that institution size and quality of research environment are unlikely to explain the differences between institutions.
Where EDI restricts speech
Principles of freedom of speech aren’t opposed to EDI per se, and nor is AFFS. However, both are opposed to forms of its implementation that have restricted freedom of speech. In recent years, such restrictive forms of EDI have contributed to a number of well-known legal and reputational disasters.
Advance HE, the sector-wide body that works to promote EDI, has recently recognised that:
Many of the challenges that [universities] have faced in relation to free speech or academic freedom arise in the implementation of policies and initiatives related to equality, diversity and inclusion.
Similar conclusions were expressed in 2024 in the Dandridge review, in which Nicola Dandridge wrote:
I also heard from staff who suggested that this legitimate desire to protect and promote principles of EDI had on occasions translated into excessive caution as to as to what could and could not be said, even when the views in question were legitimate and lawful, albeit contentious. This was associated with a potentially rigid approach that assumed that only one interpretation of principles of EDI was acceptable, making it difficult to have an open and honest discussion about how complex issues should be interpreted and applied.
This approach to EDI had the effect of… precluding legitimate debate and discussion about contentious matters.” (Paragraphs 4.8-4.9).
The AFFS report states:
EDI programmes at universities reflect a commendable desire to create open, fair communities which welcome staff and students of any race, sex, religion, belief or any other ‘protected characteristic’ under the Equality Act 2010.”
And again:
Neither AFFS, nor the principles of free speech and academic freedom, are inherently hostile to EDI which is free speech compliant, and this study is not attempting to make a case for a reduction in EDI spending.”
AFFS is concerned with particular EDI practices that have been implemented in ways that restrict freedom of speech. The report focuses on two types of EDI-related practice that restrict free speech and are highly likely to breach universities’ legal obligations.
The first is requiring applicants to academic jobs to be committed to EDI, and to provide evidence of that commitment, as a necessary condition for securing the job – what the report calls “EDI evidence requirements”. The second is requiring academic staff members, as a duty of their jobs, to promote and support EDI – “EDI support requirements”.
Why is AFFS concerned with requirements for academics to support or be committed to EDI in particular? It’s because there’s no other set of values and ideas to which adherence is widely – or even occasionally – required in order to hold an academic position in the UK.
Consider what would hypothetically be required if other ideological positions were enforced in the manner of EDI. If the values advanced under the wide EDI banner are those of a – most commonly left-wing – identity politics ideological milieu, other positions would look quite different when similarly enforced.
A traditionalist conservative requirement might demand evidence of commitment to “loyalty, tradition, and obedience” to certain institutions; a materialist left-wing position to “redistribution, solidarity, and labour”; a feminist position to “liberation, autonomy, and care”; and a Thatcherite neo-liberal position to “responsibility, realism, and discipline”. Such trios of values and ideas can be created endlessly, and to require evidence of any of them – including EDI – for appointment to academic positions is morally wrong, restrictive of freedom of speech, and highly likely to be unlawful.
What’s striking about the examples above is that almost all of them – like EDI – contain values that are, in some forms and some contexts, pretty uncontroversial, but that when required together, and as the only required values, operate as a clear mechanism of ideological selection. EDI evidence and support requirements select academics not merely on whether or not they support EDI, but also on the priority they give to different values – what they think matters most.
All of these problems are worsened by the fact that EDI is a wide and vague banner, and sometimes extends well beyond what the relevant law requires. There are elements of EDI, or ideas often promoted and enforced under the EDI banner, that are highly controversial – including views that sex is socially constructed rather than biological, and that racism in modern societies is systemic and endemic.
As the AFFS report addresses, requirements to show support for EDI risk requiring commitment to these highly contested positions. Such requirements constitute more serious and obvious restrictions of freedom of speech and breaches of universities’ legal duties, in addition to the more general free speech problems created by requiring commitment to EDI.
What we recommend
The conclusions of the AFFS report are comparatively modest. Noting that higher levels of EDI spending correlate with higher likelihood of certain compliance failures, we don’t make a case for reducing EDI spending. Rather, we recommend “seeking out and eliminating the ways in which that higher spending is leading to compliance failures”.
We suggest that universities with higher levels of EDI spending should take increased care to secure free speech, being cautious about – and taking care to avoid – possible consequent damage to free speech and compliance failures.
They should also secure a commensurate increase of spending on free speech protection, and have a dedicated member or members of staff charged with making sure that free speech and academic freedom receive attention commensurate with that received by EDI.
Finally, they should adopt a university-wide policy of “institutional neutrality”, as recommended by free speech campaigners, the Dandridge review, and already implemented by a growing number of universities.
Evidence and its critics
Early responses to the AFFS report have been notable less for their engagement with the evidence than for their avoidance of it. One article by Syra Shakir and Sean Walton, published by Wonkhe, described the report as engaging in the “bare-faced promotion of obvious falsehoods” and a “bad-faith intervention masquerading as concern for free speech” – but provided no examples or evidence for either claim, and didn’t quote a single passage from the report or discuss a single data point. Less than a fifth of the article by word count addressed the report’s content at all.
Where the article did engage, it focused wholly on “causality” – wrongly claiming that the AFFS report didn’t control for institution size or the quality of research, and giving the misleading impression that the report claimed to prove causation rather than correlation. It also ascribed to AFFS an opposition to EDI per se, despite this being directly contradicted by the text of the report, as quoted earlier in this piece.
This pattern – bold characterisation, thin engagement – is worth flagging because it illustrates a broader tendency in debates about free speech and EDI. The data and the arguments in the AFFS report are there for anyone to read and scrutinise, and they deserve scrutiny that actually grapples with what the report says rather than with what critics would prefer it to have said.
Given the subject of this article is a report from the AFFS, I think it should be made clear that William Mackesy is a co-founder of the organisation (for those who aren’t aware).
Splendid article!