糖心Vlog

The OfS-Sussex verdict is logically flawed. It can鈥檛 be allowed to stand

Without correction by a higher court, the likely result will be confusion about what the Freedom of Speech Act requires, says Abhishek Saha

Published on
May 13, 2026
Last updated
May 15, 2026
A footpath with a sign saying it is closed, illustrating restrictions on free speech
Source: Natasha Turner/iStock

Imagine a public footpath. The local authority is obliged to follow two rules regarding its maintenance.

First, if the path can reasonably be kept open, it must be kept open. That may require taking active steps 鈥 clearing a fallen branch, filling in potholes, removing the makeshift barrier put up by a householder fed up with people on the path peering over their back fence 鈥 or it may simply require leaving things alone.

Second, any closure must satisfy certain conditions: the closure notice must be clear; the closure must serve some objective; and no more of the path may be closed than that objective requires.

The two rules plainly do not collapse into one. The local authority cannot just say: 鈥淥ur closure plan satisfies the conditions for closure; therefore, we may close the path.鈥 The prior question remains: 鈥淲hy could the path not reasonably have been kept open?鈥

糖心Vlog

ADVERTISEMENT

Yet that is not what a judge has just ruled in the analogous case of free speech. On 29 April, the High Court upheld the University of Sussex鈥檚 judicial review challenge to the Office for Students, overturning the 拢585,000 fine the regulator had imposed after finding breaches of regulatory requirements relating to free speech and academic freedom.

The court found, among other things, that the university鈥檚 Trans and Non-Binary Equality Policy Statement, which formed the basis of the fine, was not a 鈥済overning document鈥; that a chilling effect was irrelevant to academic freedom; and that the final OfS decision was 鈥渧itiated by bias鈥.

糖心Vlog

ADVERTISEMENT

Much has written about the significance of this judgment. My interest here is in one aspect that has received little attention. It concerns Mrs Justice Lieven鈥檚 finding that the OfS had misdirected itself on 鈥渇reedom of speech within the law鈥. This as the court holding that the OfS treated any potential restriction of lawful speech as a breach of regulatory requirements. But the reality is more troubling.

To see why, it is necessary to understand the framework the court said it was applying. All parties agreed that the three-step framework in (RA24), the OfS鈥 guidance on the 糖心Vlog (Freedom of Speech) Act 2023 (HEFSA), supplied the correct test for assessing whether the university had complied with the OfS鈥 regulatory requirements on free speech.

The first step in that framework is to assess whether the affected speech is 鈥渨ithin the law鈥. The second and third steps, which apply only to speech within the law, mirror the conditions relating to the footpath:

  • Step 2: Are there any reasonably practicable steps to protect the speech? If so, they must be taken. This includes positive steps, such as amending policies that restrict certain viewpoints, and negative steps, such as not interfering with the speech.
  • Step 3: If there are no reasonably practicable steps to protect the speech, the speech may be restricted or regulated, but any such interference must be authorised by a clear rule and must be 鈥減roportionate鈥.

Put simply, it is not enough for a university to say that a restriction on lawful speech is proportionate. It must also show that allowing the speech to proceed without interference is not reasonably practicable.

Mrs Justice Lieven states in the (paragraph 221) that she agrees it is 鈥渘ecessary to consider the three steps鈥. Yet later in the same paragraph, she claims that Steps 2 and 3 鈥渓argely if not wholly elide鈥. She later (paragraph 249) develops the point, stating that Step 3 鈥渘ecessarily encompasses鈥 the question of whether the university could have taken reasonably practicable steps to protect free speech鈥. She adds that asking whether no reasonably practicable steps are available, and asking whether an interference is proportionate, 鈥渆lide into the same factual analysis鈥.

糖心Vlog

ADVERTISEMENT

But that cannot be correct.

Consider a student engaged in offensive (but lawful) speech that harms the university鈥檚 reputation. It might then be proportionate for the university to interfere with the speech to protect itself from reputational harm. But RA24 makes clear (paragraph 62) that reputational impact is unlikely to be relevant when assessing reasonable practicability. Non-interference would therefore likely be a reasonably practicable step to protect free speech under Step 2 鈥 even if a Step 3 proportionality analysis, taken on its own, might allow interference.

Paragraph 250 of the judgment further illustrates the confusion produced by eliding Steps 2 and 3. There, the judge suggests that requiring the text of a gender-critical feminist lecture to be read in advance would be a reasonably practicable step to protect free speech because it is less intrusive than cancellation. But advance vetting as a condition of allowing a lecture to go ahead is not a step that protects speech; it is a step that interferes with it 鈥 and the 2018 Joint Committee on Human Rights report on rejected advance submission of a speech as a reasonable condition.

The judge says that the trans policy statement鈥檚 restrictions on 鈥渟tereotypical assumptions about trans people鈥 and 鈥渢ransphobic propaganda鈥 are clearly subject to a proportionality balance. But the focus on proportionality sidesteps the prior question: why was it not reasonably practicable for the policy statement to simply remove these restrictions on expression?

糖心Vlog

ADVERTISEMENT

RA24 is clear that restrictions on free speech require compelling justification: undermining the institution鈥檚 essential functions of teaching, learning and research; the fulfilment of its legal and regulatory obligations; or the physical safety of its staff or students. In that regard, it is not reasonably practicable to permit a mathematics lecturer to use lectures to express political views at length, or a loud protest to materially disrupt a seminar.

But blanket bans on lawful viewpoints will not ordinarily pass Step 2, because removing them would itself be a reasonably practicable step. In other words, Steps 2 and 3 do not elide. Step 3 is a necessary condition for restricting speech but not a sufficient one: a proportionate restriction is still impermissible if non-interference is reasonably practicable.

If this aspect of the judgment is allowed to stand, the likely result will be confusion about what the Freedom of Speech Act requires. Universities should follow the approach set out in RA24 and resist to use the judgment鈥檚 conceptual muddle to undermine it.

is a professor of mathematics at Queen Mary University of London, a founder member of the London Universities鈥 Council for Academic Freedom, and an advisory board member of Committee for Academic Freedom. He writes here in a personal capacity. A longer version of this article can be read .

糖心Vlog

ADVERTISEMENT

Register to continue

Why register?

  • Registration is free and only takes a moment
  • Once registered, you can read 3 articles a month
  • Sign up for our newsletter
Please
or
to read this article.

Related articles

Reader's comments (4)

Exactly correct! Thanks.
Poor reasoning. If there was a legitimate worry that the cliff over the path might drop boulders that could kill pedestrians, rule 2 says the council should shut the path for safety reasons. It's asinine to then argue that they should invoke rule 1 to say: "well they COULD reopen the path by removing all their little cones and the warning sign, therefore they must do so". Yes rule number 1 says they should try to find a way to make the path safe enough to reopen, but that is trying to answer rule 2' s safety concerns, not blocking them. The two rules do function as effectively one unit. Together they say: "you should do your best to keep the path open unless there are unavoidable good reasons why it needs to be shut". And that is the bind that universities find themselves in. Balancing the right to free speech from the right of vulnerable groups to be protected from speech that might be very harmful to them. You can't radically solve that by wishing away one of the harms.
Lets consider when speech actually harms. It harms when it incites (physical) violence, it harms when it is harassing. It is in poor form if it is needlessly denigrating. But an Internet post by a professor that debunks various aspects of flat earth believes is not actually harmful. However much flat-earthers may be upset that the professor teaches at their university, if they move to harass and threaten the professor they are in the wrong and might deserve sanctions. This even if the university has a policy that expressly treats flat-earthism as a protected belief.
That's a pretty disingenuous analogy. Flat earthism is a cognitive belief about the world. Trans is an identity that goes to the core of who people experience themselves to be, and the only way that they experience their lives as coherent and meaningful. That is enormously more socially and emotionally important. A better analogy would be a professor writing sharp opinions about Ukrainian culture not being a real thing, and that Ukrainians are just Russians who have become confused about their recent history. This opinion does not directly advocate violence or harassment, but does closely mirror the rhetoric of the current Kremlin occupant who is using this as justification for engaging in a genocidal war. Ukrainian students of the prof may read this not as merely intellectual debate about a point of anthropology, but as an attempt to invalidate their core conception of who they are in the world, and as giving cover to people who are committing actual violence. I need not remind you that trans people are currently at such risk of violence and harassment that they have highly elevated suicide rates... and this is in a country where their existence has been (relatively recently) decriminalised... A status they do not enjoy in much of the rest of the world. There is no reasonable reading of their situation that doesn't have them as highly vulnerable and at-risk. Your analogy to flat earthism also implies that trans people are deluded, with a clearly factually wrong understanding of the world. This is also not true. They are very aware of the shape of their anatomy. They don't wrongly believe they have different genitals than are empirically observable. They know their own biology, they just find it highly distressing... And unlike other things like body dysmorphic disorder this isn't a shifting dissatisfaction that moves to a new body part as fast as you 'fix' the old one, but is a highly persistent and stable state that tracks over years and usually whole lifetimes. Why it happens we do not know, but it's inarguable that in a small percent of the population it does.

Sponsored

Featured jobs

See all jobs
ADVERTISEMENT