Background
The “Sullivan Review” was commissioned by Michelle Donelan (then Minister for Science) in February 2024. She announced it as a response to “the slow creep of wokeism” in science, marking it as a close cousin of the “anti-DEI” yahooism currently causing immeasurable damage to scientific research in the US.
This article is concerned with the first of the two reports this review has released. The initial report, which we will refer to as the “Sullivan Report”, was published on 19th March 2025, and is concerned broadly with “data, statistics and research on sex and gender”. In total, it makes 59 recommendations, many of which have profound implications for transgender people. Of those, 10 are picked out as “key recommendations”, and we’ll refer to these as KR-1 to KR-10.
A second report in the series was released on 2nd July 2025. It is concerned solely with complaining about criticism levelled at “gender critical” academics. We will consider the second report no further here; Nathan Oseroff has provided some detailed analysis and debunking of its claims in this Bluesky thread.
The Sullivan Report Is Not Independent
Despite repeatedly describing itself as “independent”, the report is notably not. All those involved with it have close connections to the anti-trans organization Sex Matters.
Alice Sullivan, the author of the report, is a member of the Sex Matters advisory group.
Appendix One of the report contains a legal opinion on the workability of Sullivan’s recommendations, written by Timothy Pitt-Payne, who is the husband of Naomi Cunningham, the chair of the Sex Matters board of trustees.
Providing assistance to Sullivan was Murray Blackburn Mackenzie, another anti-trans group, whose founders include Lucy Hunter Blackburn, again a Sex Matters advisor.
Further assistance was provided by Kathryn Webb who describes herself as “researching sex and gender” at Oxford University’s sociology department. Another member of that department with similar interests is Michael Biggs (aka Henry Wimbush), a trustee of Sex Matters who is frequently cited within the report.
This report is, effectively, a Sex Matters review, and it is therefore unsurprising that it is transparently a vehicle for attacking trans inclusion in society.
It is worth noting that subsequent to publication, Sullivan stated in an interview that her position in relation to trans people was one of “agreeing with Donald Trump … he is simply saying that there are two sexes and this matters”. Trump’s method of saying this is to remove all legal protections from trans people and make discrimination against them legal.
Also of interest, in terms of the authors’ stance towards trans people, is that Pitt-Payne’s “Technically Human” blog on Substack includes a post called “Digital Brains In Silicon Vats” which rehearses the conspiratorial obsessions of Jennifer Bilek, a conspiracy theorist who notoriously invokes "billionaire trans activitists" to launder antisemitism. The post, which discusses the possibility of a cadre of rich immortals eternally ruling the world, and which connects these thoughts with “transgenderism”, is effusively praised by Bilek in a comment.
The authors are clearly not disinterested parties.
Despite KR-1 of the report recommending alignment with the UKSA Inclusive Data Taskforce Review, Sullivan notably did not follow the inclusivity recommendations of that report on community engagement and trust building.
At no point before publication were trans-led organizations or the wider trans community given the opportunity to contribute to, review, or comment on the report or its recommendations, despite the far-reaching and devastating effects that implementing it would have on the trans population. (Note: the three words “trans”, “transgender”, and “non-binary” are each used more than 200 times in the report).
Sullivan’s Conception Of “Sex”
The report’s main focus is urging that it is important to gather data on a property that Sullivan variously calls “biological sex, natal sex, sex at birth”.
For clarity, we will call this “sex per unmodified certificate” (SPUC), as it conforms to the “sex” value on a birth certificate that has not been modified or replaced at a later date.
The report recommends that:
- SPUC should be the target of all questions about “sex” (KR-3)
- data about SPUC should be ubiquitously collected “by default in all research and data collection commissioned by government and quasi-governmental organisations” (KR-2)
Sullivan suggests that KR-3 “future-proofs” data collection against change because:
"Sex as a biological category is constant across time and across jurisdictions" (p5)
However, that is trivially untrue, as different jurisdictions categorize sex differently. There is no universal consensus that “binary sex” identifies natural kinds, and there is no bar to the UK updating legislation to reflect a changing scientific and social understanding of sex differentiation.
Sullivan’s vision of “future-proofing” also involves reversing the trend of the sole graph in her report (p4), by rewinding language usage back to the 1940s so that “sex” qua SPUC is used in place of concepts such as “gender” or “gender identity”.
As we will see, this would be extremely detrimental to the ability of the trans population to participate in civil society; some specific recommendations are potentially life threatening.
Sullivan’s justification is that there has been a “loss of data” about SPUC, and that this rewind would achieve ongoing “consistency” or “clarity” in data:
“The importance of clarity is a recurring theme in this report.” (p22)
This quest for “conceptual clarity” is effectively a demand that data collection and therefore services that use that data in the UK swallow “gender critical” metaphysics whole. This would require adopting a world view entirely neglectful of gender’s biosocial entanglement.
To illustrate how damaging this would be, I will first examine the serious issues that arise in the context of a specific recommendation, KR-7.
Problems With KR-7
KR-7, like several of Sullivan’s recommendations, urges that a very concrete, specific actions be taken by an individual organization:
“The NHS should cease the practice of issuing new NHS numbers and changed ‘gender’ markers to individuals, as this means that data on sex is lost…”
It’s extremely striking that a sociological review is presuming to dictate how a medical service operates, and notable that this is significantly more assertive than Recommendation 28 of the Cass Review which suggested only that the process be reviewed by appropriate organizations to address any implications.
In other words, KR-7 appears to be a startling overreach; it might be thought that Sullivan must therefore have extremely compelling evidence to support it. Ultimately, the recommendation can only be justified if it positively affects the quality of care that patients receive.
So it is that Sullivan strongly asserts that the ability to change gender markers has the result of "putting individuals at risk regarding clinical care, screening, and safeguarding".
However, the examples that Sullivan presents to support this view are preposterously bad where not completely erroneous.
Clinical Care
First of all, let’s consider clinical care.
The sole example of "sex-specific" treatment requirements that Sullivan raises (p91) relates to kidney dialysis. However, using SPUC to evaluate values for trans patients would not provide medically sound guidance, because eGFR values are affected by gender affirming hormone treatment, as per Jue et al (2021) (which Sullivan does not cite).
Of course, the reference ranges for hormone levels themselves in such patients will be those of “acquired” gender; achieving that is the intent of the treatment. If SPUC- associated reference levels are used, then routine monitoring will routinely be taken to signal alarming medical conditions.
Likewise, the same applies for reference levels with respect to many other blood tests. For example, Nolan and Cheung (2025) found that for four common blood tests, reference ranges aligning with gender identity should be used for individuals established on gender-affirming hormone therapy.
So the ability to change gender marker is important in ensuring the delivery of appropriate clinical care for many trans patients. Sullivan ignores that the “material circumstances” of trans bodies are affected by the process of transition, and cites none of the work on this issue, nor offers any specific research or evidence about harms in the patient population.
Good-faith approaches to addressing the issues that trans patients do face would include research on appropriate reference ranges or allowing the reference range to be selectable per test.
Implementing KR-7 to create a system in which gender markers are immutable rather than allowing patient choice will directly cause clinical care issues for trans patients: the exact inverse of what Sullivan claims.
Screening
Sullivan mentions screening repeatedly, and refers to the "potentially fatal consequences" (p19) of trans people not being invited to appropriate health screenings. However, the sole citation she gives on this is an article in the Evening Standard. Worse, she is amplifying the article’s headline while ignoring the content.
This article, in fact, discusses at length the fact that trans people can get access to the screenings they need (see NHS guidance on this). It also notes that medical history, as well as birth status, plays a strong role in determining what screenings are appropriate: the need for mammograms is prompted by having breasts, not by SPUC; the need for cervical screening is modulated by hysterectomy as well as trans status; and so on.
Preventing patients from changing gender marker will therefore also have “potentially fatal consequences”, and a good-faith effort to improve screening invitations for trans people and others would instead focus on solutions such as introducing anatomical inventories.
Sullivan asserts that “only males suffer testicular cancer” (p17) as an example that some diseases simply are “sex-specific” as per SPUC. But this is a generic statement that doesn’t consider that individuals suffer from those diseases only contingently, and it might be more enlightening in the context of trans healthcare to consider the universally true statement that “people who don’t have testicles can’t suffer from testicular cancer”.
Safeguarding
Sullivan explicitly says that there is “a particularly serious safeguarding risk” associated with gender marker change for under 18s, advancing as evidence a single unsourced anecdote attributed to “a paediatrician and safeguarding expert”. This tells us that a mother had requested a change to the gender marker of an infant, that the GP had complied, and that Children’s Social Care did not view this as a child protection issue. The anecdote implies that this has caused later behavioural issues, but no explanation of how they are linked is provided.
Sullivan herself said “when I first heard about that case, I was physically sick”. She expects us to likewise recoil at it, to take it as solid evidence of the anti-trans conspiracy theory that there is a cohort of parents working to forcibly “trans” their children, with the willing acquiescence of “woke” professionals.
Yet a possible alternative explanation is that it describes a situation in which the parent of an intersex child decided to change the assignment of sex made at the time of birth, in the light of new information. It is also possible to correct a birth certificate without a GRC for people who are “intersex or have a variation in sexual characteristics”.
Of course we cannot determine what happened because all we really have is a “gender critical” urban legend, a story without a concrete source, relayed by “a friend of a friend”.
It is unclear why someone conveying what is (at best) third hand information should think themselves in a stronger position to evaluate risk than the GP and social workers that were directly involved.
Sullivan’s assertion of “particularly serious” risk should be evaluated in light of it resting on a single non-verifiable anecdote for which an alternative explanation is available.
Administration and Research Effects
There are two other considerations that Sullivan puts forward in favour of KR-7, which are a) that changing a person’s gender marker “places an undue administrative burden on NHS services”; and b) that it “it undermines the integrity of data that is used to improve national clinical practice”.
Firstly, administrative costs might prompt consideration of how the system could be redesigned to make the process simpler, but a cost cutting exercise cannot be a valid reason to make changing one’s gender marker impossible. The cost of reworking the UK’s entire benefits system to accommodate gender change was addressed in §91 of the judgment in Goodwin vs The UK in the following terms:
“the Court considers that society may reasonably be expected to tolerate a certain inconvenience to enable individuals to live in dignity and worth in accordance with the sexual identity chosen by them at great personal cost.”
Secondly, Sullivan’s notion of riding roughshod over patient identity merely for the sake of the convenience of researchers is not only ethically unjustifiable but also is a recipe for poor research outcomes. The report ignores the literature about research design which takes account of the entanglement of sex with gender, and it also entirely neglects the impact of gender on the effectiveness of healthcare interventions. We’ve already seen that bad assumptions based on SPUC can lead to incorrect diagnosis and treatment; those same assumptions should not be embedded in research practice.
Likely Effects Of Implementing KR-7 On Patients
Given these flimsy and erroneous rationales, it is utterly reprehensible that the Prime Minister’s office immediately said that this review’s recommendations represented “just common sense” for healthcare immediately following the release of the report, and that Wes Streeting has already moved to implement KR-7 for under 18s.
Worse, it appears that at no point did Sullivan or the Health Minister consider any of the other potential impacts of implementing it. There is no record of the affected population, patient advocacy groups, or experts on transgender health being consulted on KR-7 before the recommendation was made or accepted.
It is therefore unsurprising that the review fails to evaluate the catastrophic impact that KR-7 would have in eroding the quality of patient care across every interaction that trans patients have with the NHS, driving disengagement from healthcare services. Systems that disrespect patient identity and that misrepresent medical history are known to be a major barrier to trans people accessing healthcare of all kinds.
By forcibly outing patients as trans (throughout their entire lifetime, including to non-medical staff such as receptionists), KR-7 would cause every medical interaction to become a potential source of prejudicial treatment, distress, and/or misdiagnosis through the phenomenon of “trans broken arm syndrome”. Doing so would also grossly infringe on trans patients’ basic right to a private life, by making personal history apparent to others when medically unnecessary to do so.
Given these major failings, and the dangerous consequences of this recommendation, implementation of KR-7 should be revoked immediately, and any further implementation of any of Sullivan’s other recommendations completely halted pending a full analysis of both their rationales and their likely wider impacts.
Legal Questions Relating To The Report
Beyond the specific issues that individual recommendations bring (and there are many issues across the 59), there are fundamental questions about the soundness of the report as a whole, due to its core positions relying on very questionable legal assertions.
The Report Is Built On A False Dichotomy
The stated aims of the report presuppose that “sex and gender identity” are separate properties that must be recorded separately.
This is also explicitly asserted at several points, for example:
“Sex and gender identity are distinct characteristics and therefore should not be treated as substitutes.” (p23)
Much ink is spilled within the report on trying to convince the reader that “sex” and “gender identity” are non-overlapping, despite the fact that “sex” and “gender” are frequently used as synonyms, including in legal use:
“GRA 2004 uses the terms sex and gender interchangeably. It is reasonably clear that in GRA 2004 ‘gender’ is being used as a synonym for ‘sex’.” (p22)
All this is intended to direct our attention away from the simple fact that the intended or appropriate referent of “sex” is often “gender identity”, which represents “social” or “lived” sex, rather than SPUC.
A good example of this is recommendation 40, which suggests renaming the gender pay gap to the “sex pay gap”, despite Sullivan’s own citation discussing that much of the gap is driven by social rather than biological factors, and despite research that suggests trans women are subject to a much larger pay gap than women in general.
(The suggested term “sex pay gap” also inadvertently exposes the infelicities that would be attendant on a search-and-replace of “sex” for “gender”; the actual implementation of such terminology manipulation in scientific data by the Trump administration, and the problems this poses are discussed by Freilich, Janet et al.).
The report aims to persuade its audience that the false dichotomy between “sex” and “gender identity” is recognized in law, as per the assertion that treating them equivalently might constitute a legal breach:
“Recording data on sex as gender identity is likely to be a breach of UK GDPR and the DPA 2018, read together, and likely to breach article 8 ECHR.” (p44)
This is Timothy Pitt-Payne’s legal opinion, but if that opinion was to be upheld in court it would cause UK GDPR to sharply diverge from EU GDPR regulations; six days before the Sullivan Report was released the Court of Justice of the Europen Union (CJEU) ruled that the right to rectify inaccurate data includes replacing SPUC with gender identity.
The CJEU ruling follows the common sense principle that when the purpose of recording data is identifying an individual, the appropriate target is their identity, and not a description assigned at birth that they actively repudiate.
It is of course possible that the UK will diverge from this view, and ultimately undermine data adequacy in order to cement “gender critical” metaphysics into law, but that divergence would potentially be hugely consequential for the UK as a whole.
Asserting there is a legal difference here would also require the UK rejecting the entire body of rulings from the European Court of Human Rights (ECtHR) on “gender identity issues”, which has repeatedly found Article 8 violations over decades in respect of enforcing the recording of sex as SPUC.
While some “gender critical” voices advocate leaving the European Convention on Human Rights, torching the entire roster of human rights in the UK seems like an extremist position, even today.
These GDPR and ECtHR rulings entirely contradict the Appendix One reasoning in §§86- 91 that the review relies on; therefore this central plank of the report is entirely rotten.
The repeated ECtHR rulings on privacy also contradict the absurd Appendix One assertion of “we can always tell” as a legal principle:
“If a person’s sex at birth is readily apparent to anyone who encounters them (e.g. from physical appearance, or voice) then it is hard to see how there can be any reasonable expectation of privacy in the mere fact that their sex at birth is male or female.”
The ECtHR rulings were not contingent on assessments of the physical attributes of a person, or any other form of transvestigation. Indeed, the very opposite is true: Article 8 violations were found to have resulted from the recognition of identity being made contingent on medical intervention (for example, X and Y v. Romania).
Key Recommendations Are Unworkable
SPUC is not information that trans people will (or should) provide routinely.
Asking a question about it can, unless handled carefully, read as a demand to deny one’s personal identity; answering it is also likely to reveal trans status and therefore carries with it the threat of (potentially ongoing) discrimination, harassment, or mistreatment related to incorrect assumptions about the social or physical realities of one’s existence.
Unless there are powerful explanations of the relevance of this data and guarantees that it will not be misused, asking for SPUC will therefore generally be met with refusal or disengagement. Mandating its use ubiquitously therefore directly militates against the possibility of achieving “data accuracy” while also directly creating barriers to trans people fully participating in society. (Note also with respect to the ubiquity of collection that Sullivan urges, GDPR data minimization principles do not allow for harvesting data about gender in contexts where it has no relevance).
Such issues are obviously part of the reason that “gender” or “gender identity” is today widely used rather than asking for “sex” qua SPUC.
The report attempts to reverse this move away from the terminology of “sex” by asserting that “gender identity” should not be used as a question target (see KR-5 and KR-6); it supports this by advancing the hypothesis that there is a category of people who “do not understand themselves as having a gender identity at all” (see Appendix One §35, §88, and §§97-98).
Again, legal peril is threatened:
“Holding gender identity data about a person who does not have a gender identity is likely to be a breach of data protection legislation” (p44)
This is fatuous for the following reasons:
- The category of people that Pitt-Payne is discussing are people with “gender critical” beliefs. These people have a gender identity, whether they “believe” they do or not, because gender identity incorporates one’s understanding of one’s sex. This is said in §35 to be “contentious”, but we’ve already seen that it is erroneous to hold that “sex” and “gender identity” are entirely distinct in law.
- The “belief” of “not having a gender identity” in these terms is akin to not understanding oneself as having a name. One is inevitably lumbered with such a characteristic in virtue of a “naming ceremony”, such as the status being recorded on a birth certificate (or through a later declaration of an alternative).
- The notion of §35 that some people might be members of such a group “because the concept is unfamiliar to them and not understood” does not mean it would be impossible for them to grasp what the concept refers to when explained. If this were grounds for a “breach” then it would potentially be a “breach” to hold any kind of personal data whatsoever.
- The identity of genuinely agender people can be accommodated by an “agender” option or by making the question optional. But “gender critical” people do not claim such an identity: they want to be identified as a “woman” or a “man”.
Pitt-Payne’s argument represents the attempted weaponization of an employment tribunal ruling which stated that gender critical beliefs concerning gender identity were protected.
However, the right to hold a belief does not imply that the belief is true. The ruling noted this distinction between “is” and “ought” with respect to a parallel situation:
“the legal recognition of civil partnerships does not negate the right of a person to believe that marriage should only apply to heterosexual couples”
The Report Fails To Account For Human Rights Duties
Sullivan states in KR-5 that for purposes of meeting the Public Sector Equality Duty (PSED), “sex” rather than “gender” should be used; an addendum to the report claims the recent UK Supreme Court judgment (FWS vs Scottish Ministers) vindicates this position.
The judgment does indeed discuss PSED in §§237-244, in the context of positive action measures for women, the topic of the case. For allowable exceptions to sex discrimination provisions, the Supreme Court’s definition of “sex” must now be considered, but as explicitly called out in §250, trans people can still access protections against sex discrimination on the basis of perceived sex; as per §§254-256, the same applies with respect to harassment.
The judgment is distinctly odd. It results in a situation which, in common sense terms, is clearly unfair and absurd: two women with identical careers, both treated identically in the course of their career because they are both perceived as women, must be treated differently if they are distinct solely in terms of so-called “biological sex”. This is a characteristic not mentioned anywhere in relevant legislation; the nature of it was held to be ‘self-explanatory’ by the court but, ironically, it may not correlate with any perceptible aspect of adult biology, let alone with the social realities of “sex”, in its broader sense as “gender” (which is how it is interpreted in all legislation other than, now, the Equality Act 2010).
Likewise, the judgment explicitly suggests that the members of a couple may be treated in law as being of the same sex for the purposes of matrimony but as being of the opposite “sex” for the purposes of association. This is a fundamentally incoherent position, with chaotic implications which will ultimately require further legislation or litigation to resolve.
As Crash Wigley has strongly argued, the court failed to properly analyze the human rights implications of their judgment, and the same is true of Sullivan moving in lockstep with it. As a result, both the judgment and the report directly mandate what is an unsustainable situation as per the §90 of the EctHR judgment in Goodwin v The UK:
“In the twenty first century the right of transsexuals to personal development and to physical and moral security in the full sense enjoyed by others in society cannot be regarded as a matter of controversy requiring the lapse of time to cast clearer light on the issues involved. In short, the unsatisfactory situation in which post-operative transsexuals live in an intermediate zone as not quite one gender or the other is no longer sustainable.”
The question now is to what extent the fundamental rights of trans people will be violated in the UK if they are further forced into such an “intermediate zone”.
While it is hard to foresee how this Gordian knot will be undone, the judgment may well provide additional impetus towards providing services in a gender-neutral way; so-called “biological sex” is such a rarefied abstraction that segregating people on this basis seems a very uncertain basis for a “proportionate means to achieving a legitimate aim”.
Meanwhile, mechanically following Sullivan’s recommendations simply cannot align with the PSED aims broadly. One cannot, for example, “foster good relations between people who share a protected characteristic and people who do not share it” or “advance equality of opportunity” by using an approach that is likely to cause the harassment of trans people due to violation of their Article 8 rights. Nor does using terminology that is perceived as directly hostile to specific identities fulfill this duty.
Far from guiding us to compliance with PSED, Sullivan pushes us towards breaching it.
Conclusion
The Sullivan Report is fundamentally a bloated example of the typical four-step “gender critical” pattern of argument, each step of which we can tag with phrases repeatedly used:
- assert that sex and gender are non-overlapping (“mixed/muddled target”)
- assert that gender is in some sense unreal (“conceptual clarity”)
- assert that sex is immutable (“data accuracy”)
- deny recognition of trans identities on this basis (“loss of data on sex”)
The means of achieving the desired trans-exclusionary outcome is to take us back to the conceptual landscape of the 1940s: forcing the ubiquitous use of “sex” (meaning SPUC) while discouraging the use of “gender identity” by asserting that it is illegal to hurt the feelings of “gender critical” conscientious objectors.
Prescribing the verbal usages of the past will certainly not provide genuine “conceptual clarity”, because the end result will be terminology that is not be aligned with today’s social realities (such as the existence of a class of non-binary people). It is the “clarity” brought by wearing blinkers.
Worse, as we’ve seen in evaluating KR-7, these recommendations don’t just work to eliminate the presence of trans people in data, they will also directly work to harm the trans population through forms of social exclusion. Even this most plausibly “biological” of the report’s recommendations is incredibly ill-evidenced and dangerous; the other examples of the importance of “data about sex” that Sullivan touches on are much less plausibly
SPUC-driven. Sullivan appears to think that “biosocial” refers solely to the effect of the biological upon the social, but it represents a two-way entanglement that the existence of trans people is a vivid reminder of.
The fact that aspects of the report were acted upon immediately on its release is extremely worrying, suggesting there is a realistic possibility that these recommendations may become widely adopted or widely influential, as happened with the internationally condemned Cass Review. We can already that see lobbying by Sex Matters **has worked to successfully influence ONS guidance, which often cascades to other organizations; this is picked up in this report and used as though it provides independent justification for Sullivan’s recommendations (via KR-9 and KR-10).
The UK establishment, including political and judicial bodies, have been repeatedly credulous with respect to this repeated pattern of lobbying by ideological extremists, including the use of “independent” reviews to launder anti-trans attacks on the participation of trans people in society.
We urge a future robust stance of profound scepticism as to the influence of damaging “anti-DEI” ideology within future reports, rather than an eager welcome to it as “common sense”. The badly evidenced, directly harmful recommendations of both the Sullivan and Cass reviews should have been rejected entirely.