A long-form commentary on the Statutory Code of Practice for Services, Public Functions and Associations, published by the Equality and Human Rights Commission on 6 May 2026.
Some regulatory documents announce themselves on the first page. Others reveal themselves only on the second reading. The Equality and Human Rights Commission’s Statutory Code of Practice: Services, Public Functions and Associations, laid before Parliament on 6 May 2026, takes a third route. It discloses what it is doing in worked examples buried two hundred pages in.
What it is doing, on a careful reading of its 342 pages, is this. It endorses, in worked-example form, the lawful exclusion of a trans man from a women’s domestic violence counselling group on the ground that his appearance distresses other survivors. It invites every gym attendant, librarian and village hall trustee in the country to approach anyone whose “physique or physical appearance” is judged ambiguous and require them to confirm their sex by means the Code declines to specify. It deputises lay service providers to perform a proportionality assessment of formidable legal complexity at the front door, in real time, under the threat of litigation from two directions simultaneously. And it does all of this while announcing itself as the faithful implementation of a Supreme Court judgment that, on any honest reading, decided nothing of the kind.
The Code is issued under section 14 of the Equality Act 2006. By virtue of section 15 of that Act, courts and tribunals must take it into account where it appears relevant to any question before them. It will govern the conduct of every refuge, hospital, sports club, gym, café, library, school and public toilet in England, Scotland and Wales. It is the most consequential piece of equality-law guidance to appear in this jurisdiction in more than a decade.
It is also a document that performs equality while delivering its opposite.
The Code is internally contradictory. It strips the Gender Recognition Act 2004 of meaningful effect for the purposes of the central piece of UK anti-discrimination legislation, and then declines to acknowledge that this is what it has done. It addresses the problem of how a service provider is to determine a person’s biological sex by largely refusing to address it. It treats the Supreme Court’s narrow holding on statutory construction as if the Court had handed it a mandate to rewrite the architecture of equality law from the ground up. It has not.
What follows is a long-form commentary on what the Code says, what it leaves out, what it gets wrong, and what its publication reveals about the institutional condition of the body that produced it. The argument is straightforward. The Supreme Court delivered a narrow ruling on the meaning of a word. The EHRC was under no obligation to convert that ruling into a sprawling regulatory framework. It chose to. The framework it produced will protect no one cleanly, satisfy no one fully, and multiply the litigation it claims to clarify. It is a Code that will be cited for years, in every direction, by claimants whose interests it purports to serve and whose lives it has, in fact, made more difficult.
What For Women Scotland actually decided
The Supreme Court’s decision in For Women Scotland Ltd v The Scottish Ministers [2025] UKSC 16 was handed down in April 2025. It concerned the proper construction of the word “woman” in the Equality Act 2010, as deployed in the Scottish Government’s statutory guidance on the Gender Representation on Public Boards (Scotland) Act 2018. The campaigning organisation For Women Scotland argued that the Scottish guidance had impermissibly extended the meaning of “woman” beyond what the 2010 Act would bear.
The Court agreed. It held that “sex” in the Equality Act 2010 means biological sex, and that a Gender Recognition Certificate issued under the Gender Recognition Act 2004 does not change a person’s sex for the purposes of the 2010 Act. It grounded that holding in the statutory text, the structure of section 4, and the practical workability of the various exceptions in the Act’s schedules.
That is what the Court decided. It is essential to be clear about what it did not.
The Court did not abolish the protected characteristic of gender reassignment, which remains in section 7 of the Equality Act 2010 in unaltered terms. It did not invalidate the Gender Recognition Act 2004, which remains on the statute book. It did not opine on the operation of single-sex services, hospital wards, communal accommodation, sport, toilets, refuges or prisons. It did not address how a service provider is to determine a person’s biological sex in practice. It did not consider, in any depth, the implications for Article 8 of the European Convention on Human Rights. The judgment is rigorous and narrow. It does what apex courts in common-law systems are supposed to do: it resolves the legal question presented and leaves the regulatory consequences to be worked out by Parliament and the relevant statutory bodies.
That is the discipline that an apex court is meant to display. It is also the discipline that the EHRC, in answer to the Court, has comprehensively abandoned.
From narrow judgment to sprawling Code
The EHRC could have responded to For Women Scotland with a short technical revision. A few amended paragraphs in the existing Code, confirming the meaning of “sex”, updating the worked examples, and otherwise leaving the architecture of equality law intact. That would have been a modest, faithful, defensible response to a narrow judgment.
The EHRC did not do that. It produced, instead, a 342-page document that pursues the For Women Scotland reasoning into every domain of public life to which the Equality Act applies. Toilets, changing rooms, hospital wards, refuges, sports clubs, communal dormitories, insurance underwriting, schools, religious associations, charities, political parties. Every category of service, in every direction, with worked examples whose drafting frequently betrays the absence of meaningful contact with the user groups the examples purport to describe.
This was a discretionary choice. It was not required by the Supreme Court, which was meticulous about the limits of its own holding. It was not required by Parliament, which has not, since the For Women Scotland judgment, legislated on any of the matters the Code now purports to settle. It was not required by the Government, which retains the power under section 14(7) of the 2006 Act to direct the EHRC to produce a Code on a specified subject but did not do so. The Commission decided, on its own initiative, to convert a narrow point of statutory construction into a regulatory programme.
In choosing this course, the EHRC has placed itself in the position of having to defend the consequences. Those consequences, as the following sections show, are not defensible. They are not defensible because the framework the Code installs is internally contradictory; because the proportionality test it deputises is unworkable at the front line; because the verification procedure it sketches is an invitation to discrimination; and because the people whose interests the Code claims to advance — women, trans people, service providers, ordinary members of the public — are, on any sober reading of what the Code in fact says, all left worse off.
The trans man paradox
Begin with a real person. He is twenty-six. He was assigned female at birth. For the past four years he has lived as a man. He has had top surgery, takes testosterone, has a beard, and his voice is, by now, indistinguishable from any other man’s voice. He has a Gender Recognition Certificate, obtained at no small administrative effort. Three weeks ago, his former partner, a woman, attacked him with a knife in their kitchen. He survived. He is now seeking the support of the only domestic violence counselling service in his town.
Under the Code, he is a woman. The Code says so explicitly at paragraph 2.50: “a trans man with a GRC is a woman… for the purposes of the Act.” He therefore has, on the EHRC’s own framing, a statutory entitlement to use women-only services. He also retains, in unaltered form, the protected characteristic of gender reassignment under section 7 of the Equality Act 2010.
So far, so coherent. Two minutes’ further reading destroys the coherence.
At paragraph 13.146, the Code provides that he “might be excluded from the women-only service if the service provider decides that, because he presents as a man, other service users could reasonably object to his presence, and excluding him is a proportionate means of achieving a legitimate aim.” At paragraph 13.150, the Code offers a worked example. The setting: “Group counselling sessions are provided for female survivors of domestic violence.” The decision: “The service provider excludes a trans man from the sessions because they consider he is likely to be perceived as a man and the service provider is concerned that women service users could reasonably be worried or distressed by the presence of someone they perceive to be a man using the service.” The Code’s conclusion: this exclusion is likely to be lawful.
Stop and consider what has just been said. The trans man at the door is, in law, a woman. He has, in law, a right to use the service. He has the protected characteristic of gender reassignment. The reason advanced for his exclusion — that he “presents as a man” — is precisely that protected characteristic. The justification advanced for overriding his statutory entitlement is the discomfort of third parties.
He cannot, of course, go to the men’s service. There is no men’s domestic violence service in his town. There is barely a men’s domestic violence service in the country. He is also not, in law, a man. He is suspended between two categories that the Code is content to admit do not, between them, accommodate him.
The Code, to its limited credit, briefly acknowledges this risk. At paragraph 13.148: “in the case of services which are necessary for everybody, such as toilets, it is very unlikely to be proportionate to put a trans person in a position where there is no service that they are allowed to use.” Having made this admission, the Code averts its eyes and carries on. It does not explain what happens to the man at the door. It does not consider whether the counselling service is, for him, “necessary.” It does not return to him at all.
The drafters have produced a worked example in which the country’s equality regulator instructs a domestic violence service to turn away a survivor of domestic violence on the ground that his face distresses other survivors. They have done this, apparently, without noticing that this is what they have done. It is not a fringe possibility extracted from the Code by an unfriendly reader. It is the conclusion the Code itself reaches, in language the Code itself supplies, on facts the Code itself constructs.
This is paragraph 13.150. There are 341 other pages.
The mirror problem: trans women
The picture for trans women is, structurally, the mirror image of the picture just described, with an additional twist that makes it worse.
A trans woman is, for the purposes of the Equality Act, a man. The Code so provides at paragraph 2.50: “a trans woman with a GRC is a man… for the purposes of the Act.” She retains the protected characteristic of gender reassignment. The same proportionality framework at paragraph 13.146 permits her exclusion from women-only services on the symmetrical reasoning: she is, in law, a man, and the Schedule 3, paragraph 28 exception applies.
The twist is at the men’s service. The Code’s logic, applied evenly, must permit the exclusion of a visibly female-presenting trans woman from a men’s facility on the same “reasonable objection” reasoning that was used to exclude the trans man from the women’s facility. The structural argument is identical: the discomfort of third parties overrides the statutory entitlement of the excluded person.
The trans woman is therefore suspended between categories on exactly the same logic as the trans man. She is legally a man. She cannot use women’s services. She is visibly a woman and can be excluded from men’s services on the same proportionality reasoning.
The Code, however, contains no worked example of a trans woman excluded from a men’s service. There are no such examples anywhere in the 342 pages. The Code’s framework permits the exclusion in both directions; it provides the worked example only in one. Whether this is editorial cowardice or editorial preference is not a question the Code allows the reader to answer.
What the Code in any event achieves, in the operation of its own framework, is the production of a class of people — trans men, trans women, in either direction — whose statutory entitlement to single-sex services is contingent on the comfort levels of the people who get there first. That is not equality law. It is a queue.
The hospital ward, the toilet, the changing room
Consider the same person across three services on three consecutive days.
On Monday, he goes to an obstetrics and gynaecology outpatient clinic, because biology is biology and the clinic exists to serve people with the anatomy he has. The Code’s worked example at paragraph 13.153 explicitly contemplates this. The OBGYN service is one “only women and trans men need to use.” Where the service preserves dignity and privacy, the Code states, “it is unlikely to be proportionate to exclude a trans man because of objections from female service users.” He is admitted. He is treated. He goes home.
On Tuesday, he tries to use a public toilet at the railway station. The Code endorses, at paragraph 11.887, a sports centre that provides separate-sex toilets alongside “an accessible toilet… so it can also be used as a mixed-sex toilet for anybody who does not wish to use the toilet for their sex.” The railway station has no such facility. The station has men’s toilets and women’s toilets and nothing else. The Code does not address what is supposed to happen to him at the railway station. He goes to the women’s toilets, because that is what he is in law. The cleaner objects. The Code, at paragraph 13.168, provides that the cleaner may “ask individuals to provide confirmation that they are of the eligible sex by proportionate means” where there is “clear evidence of an issue with members of the opposite sex… seeking to access” the facility, including “the individual’s physique or physical appearance.” He has the physique and physical appearance of a man. The cleaner asks. He has no documentation on him.
On Wednesday, he attends the domestic violence counselling group. He is excluded.
The same person, in the same week, in the same body, is in turn welcomed for clinical reasons, interrogated for cosmetic reasons, and turned away on the discomfort of third parties. The Code, applied as written, produces all three outcomes. The drafters appear to think that this is consistent. It is not consistent. It is the random output of a framework whose decision rule depends on factors that vary by setting and by audience and by the prevailing temperature of opinion among whichever members of the public are present that morning.
The Code’s defenders will protest that paragraph 13.153, the OBGYN example, demonstrates the framework’s flexibility. They are wrong about that. What it demonstrates is that the framework can produce a humane outcome where the service provider is a regulated clinical body acting on biological reality, and an inhumane outcome where the service provider is an unregulated lay body acting on the preferences of those who got there first. The framework does not distinguish between these cases on principled grounds. It distinguishes between them on the practical question of whether the third-party “reasonable objection” was sufficiently loud at the door.
That is not a framework. It is a barometer.
Domestic abuse services: where the stakes are highest
The Code’s treatment of domestic abuse services deserves separate attention because the stakes for the people affected are higher than in any other category of service the Code addresses, and because the consequences of getting it wrong are uniquely difficult to undo.
Refuges and counselling services exist because the people who use them are, in many cases, in immediate or ongoing physical danger. The decisions made at the point of admission to such a service are decisions made at the point of greatest vulnerability in the lives of their users. They are also decisions made by service providers who are typically small, charitably-funded, legally exposed and acutely risk-averse.
The Code’s framework lands on these providers a proportionality assessment of formidable complexity. They are to consider, on a case-by-case basis, whether the inclusion or exclusion of any particular trans person is a proportionate means of achieving a legitimate aim. They are to weigh the dignity and safety interests of the female survivors already in the refuge. They are to weigh the dignity, safety and statutory entitlement of the trans person at the door. They are to consider whether there is a suitable alternative service available — and where there is not, the Code admits that exclusion will not be proportionate, though it leaves the consequences of that admission unworked out.
In practice, the framework will produce nothing of the kind. The small refuge, with three caseworkers and a waiting list of women in present physical danger, will not be in a position to conduct a bespoke proportionality assessment of every applicant against every other applicant in the building. It will adopt a written policy. The written policy will exclude trans women, or include them, and either policy will produce litigation against the refuge from one direction or the other. Faced with that prospect, the rational response of the sector is to exclude — because exclusion produces a smaller and more easily defended class of claimants, and because the Code itself, in its worked examples, has already provided the legal cover.
The same logic operates against trans men, with the additional injustice that the parallel men’s services to which they are nominally entitled do not, in practice, exist. The Code’s worked example at paragraph 13.150 will not, in any real refuge, be the rare and tightly-justified exception its drafters appear to imagine. It will be the template.
This is the operational consequence of the Code. It is not what its drafters would presumably claim to want. It is, nonetheless, what the framework, applied as written, by sober small charities acting on legal advice, will produce. The Code has built, into the foundations of its domestic abuse framework, an incentive to exclude. The incentive will be heeded. The people who will discover this, in the years to come, are the trans people fleeing violence, who will find that the equality regulator has, in their name and for their protection, made the door harder to open.
Sport: special carve-out, special confusion
The Code’s treatment of competitive sport is the case study in regulatory delegation. Sport sits under a special statutory exception in section 195 of the Equality Act 2010, which permits sex-based restrictions in “gender-affected activities.” The Code addresses the topic at paragraphs 13.65 to 13.78.
At paragraph 13.73, the Code states that sex-based rules in gender-affected activities “should be applied on the basis of biological sex,” and that “trans people should not be included in single-sex or separate-sex competitions for the sex with which they identify.” Three lines later, the Code admits that “the law on the exception for sex discrimination in relation to gender-affected activities in section 195, paragraph 1 is not settled.”
The Code therefore instructs sports providers in the most directive language available — “should” — while admitting, in the same paragraph, that the legal foundation for the instruction is unsettled. There is no clearer demonstration of the Code’s regulatory ambition outrunning its legal warrant.
The example at paragraph 13.74 is the giveaway. An athletics club organising a women’s running event must consider: indirect sex discrimination from women if it includes trans women; direct sex discrimination from men if it excludes men but includes trans women; direct gender reassignment discrimination from trans women if it excludes them. The Code helpfully points out that whatever the club decides, it is exposed to litigation from at least one direction and possibly from three.
The Code’s answer is that the club must reason its way through this on a “case-by-case basis.” This is not guidance. This is a shrug. Section 195 is one of the most technically demanding provisions in the Equality Act. Its application to trans athletes engages international sporting regulations, the medical evidence on hormone treatment, contested empirical questions about athletic performance, and safety considerations that vary by sport. The proposition that an amateur athletics club with a treasurer who keeps the accounts in a notebook is in a position to make a robust and litigation-proof proportionality assessment on these factors, for each athlete, in each event, is not a regulatory proposition. It is regulatory theatre.
What this part of the Code will in fact produce is a wave of cautious blanket exclusions — not because amateur sports clubs are hostile to trans people, but because exclusion is the only outcome that closes off the most exposed litigation risk. The Code will, in this respect, have achieved the opposite of its stated purpose: it will have driven trans people out of amateur sport altogether, not through any decision Parliament took, but through the operation of a framework that no service provider in the country has the legal resources to apply.
The verification question
The deepest silence in the Code is its treatment of the question it cannot avoid: how, in practice, is a service provider to determine a person’s biological sex?
The Code addresses this at paragraphs 13.160 to 13.182, in a section headed “Asking about protected characteristics.” Paragraph 13.161 acknowledges that information about a person’s sex is personal data and “may also be protected under Article 8 of the European Convention on Human Rights.” Paragraph 13.162 states that the request must be “objectively justified” to comply with Article 8. Paragraph 13.165 warns that a request made in a manner that is “rude, combative or offensive” may itself constitute harassment. These are the safeguards.
Paragraph 13.168 is where the safeguards collapse. The Code provides that “where, regardless of such communications, there is clear evidence of an issue with members of the opposite sex accessing or seeking to access the single or separate-sex service or association in question, it may be legitimate to ask individuals to provide confirmation that they are of the eligible sex by proportionate means. Evidence of such concern might include the individual’s physique or physical appearance, behaviour or concerns raised by other service users.”
Read that paragraph slowly. The Code is telling every service provider in the country that they may approach a person whose “physique or physical appearance” is judged inconsistent with the eligible sex of the service, and require them to confirm their sex by means the Code declines to specify. The triggers — physique, appearance, behaviour, “concerns raised by other service users” — admit no objective test. They are, in every operative respect, the kind of subjective triggers that produced the worst excesses of the historical policing of public space along racial lines. The Code does not appear to have noticed the analogy.
The class of people who will be subjected to this interrogation includes not only trans people but butch lesbians, gender non-conforming women, tall women, women with deep voices, men with delicate features, teenagers whose development has not yet aligned with adult sexed norms, and anyone else whose body or comportment does not conform to the operator’s idea of what a member of the eligible sex looks like. The Code’s only safeguards are the requirement of proportionality and the warning against doing it “rudely.” There is no provision for what should happen if the person refuses to answer. There is no provision for what counts as adequate confirmation. There is no provision for the data protection consequences of recording the answer. There is no provision for whether the operator may detain, eject, or call the police on a person who declines to engage with the question.
A serious Code would have set out a verification procedure: what may be asked, in what circumstances, by whom, with what documentation, and with what consequences for refusal. This Code does none of those things. It has produced, instead, a regulatory licence for the casual public interrogation of every visibly non-conforming person in the country, in any single-sex setting, at the discretion of anyone wearing the operator’s lanyard. The Code calls this guidance. It is, in plainer language, a permission slip.
Deputised proportionality: regulation by shopkeeper
Step back from any particular service category and look at the Code’s structural design. The same mechanism animates almost every difficult passage. The Code, faced with a difficult case, instructs the service provider to determine whether the action under consideration is “a proportionate means of achieving a legitimate aim.”
That formulation is drawn directly from the statute. Proportionality is the test in section 19(2)(d) for indirect discrimination, in paragraph 28 of Schedule 3 for the gender reassignment exception in single-sex services, in section 195 for sport, and in numerous other places. The Code did not invent the test. It inherits it from Parliament.
But the proportionality test in its statutory incarnation has been the object of forty years of appellate litigation in the United Kingdom and in Strasbourg. The leading cases on its application — and they are leading cases precisely because the application of the test is so difficult — work through factor after factor, weigh evidence at length, and reach their conclusions only after sustained judicial reasoning with the benefit of skilled adversarial submissions. The test is a high-quality device for appellate courts engaged in close reasoning. It is a calamitously bad device for front-line decision-making by lay administrators acting at speed under commercial and reputational pressure.
The Code does not appear to have grasped this. Or, if it has grasped it, it does not address it. The proportionality test is transplanted, page after page, from the appellate courts to the front line. The village hall trustee, the leisure centre duty manager, the refuge caseworker, the school caretaker, the librarian, the bouncer — every one of them, the Code blandly tells us, must in each individual case form a view as to what is a proportionate means of achieving a legitimate aim. The Code offers, by way of guidance on how to do this, worked examples whose conclusions are themselves not infrequently contestable.
There are two predictable outcomes from this design. The first is unevenness. Different providers, applying the same Code to the same facts, will reach different conclusions about what is proportionate. The decisions will not be consistent in any direction. Some will be challenged in litigation; most will not. The result will be a patchwork of inconsistent decisions across the country, against which any individual claim of discrimination is harder to bring, harder to defend, and slower to resolve.
The second predictable outcome is risk-averse over-exclusion. Faced with a choice between a defensible blanket policy and a contestable case-by-case judgment, the rational service provider will adopt a blanket policy. The blanket policy will tend to exclusion rather than inclusion, because the exclusion of trans people produces a narrower and more easily defended class of potential claimants than the inclusion of trans people in single-sex services. The Code’s drafters appear to believe that the proportionality test will moderate this tendency. It will not. It will, on the contrary, supply the legal cover that makes blanket exclusion the path of least resistance.
This is regulation by shopkeeper. It is the systematic outsourcing, by the country’s equality regulator, of the most difficult judgments in discrimination law to people whose training, time and resources are not equal to the task. The shopkeeper is not the villain. The shopkeeper is the casualty.
The GRC holder erased
The Gender Recognition Act 2004 was enacted to give legal effect to the European Court of Human Rights’ decision in Goodwin v United Kingdom (2002) 35 EHRR 18, in which the Strasbourg Court held that the United Kingdom’s refusal to recognise the acquired gender of trans people violated Article 8 of the European Convention. The 2004 Act established a quasi-judicial procedure under which a trans person, having satisfied a Gender Recognition Panel that they had lived in the acquired gender for the requisite period and met the statutory conditions, could obtain a Gender Recognition Certificate. Section 9 of the Act provides that, on the issue of the Certificate, the acquired gender “becomes for all purposes the acquired gender.”
The phrase “for all purposes” is unambiguous on its face. Section 9(3) permits other enactments to disapply the effect of the Certificate in particular contexts, but the default established by Parliament in 2004 was that a Certificate would have full legal effect across the statute book.
Twenty-two years later, the Supreme Court has held that the Equality Act 2010 — not some peripheral statute, but the central piece of UK anti-discrimination legislation — is, in effect, not within “all purposes.” The Code now implements that reading, across hundreds of pages of guidance, on every category of service to which the 2010 Act applies. A Gender Recognition Certificate, in respect of the country’s principal equality statute, is now meaningless.
The Code’s only direct engagement with the constitutional implications of this is in a single sentence at paragraph 2.51: “This Code only relates to the application of the Equality Act 2010. It does not deal with the implications of the For Women Scotland judgment for the application of section 9 of the Gender Recognition Act 2004 to any other legislation.”
That sentence is a study in evasion. Parliament enacted a statute whose central operative provision was a section establishing legal effect “for all purposes.” The country’s highest court has now held that the central UK anti-discrimination statute is outside the scope of those purposes. The country’s equality regulator, asked what this means for the other purposes Parliament had in mind, declines to engage. The implications for pension entitlements, for marriage and civil partnership registration, for the operation of EU-derived employment regulations that remain part of UK law, for the application of immigration rules to GRC holders, for the gendered structure of the criminal justice system — none of these are addressed. The Code’s drafters have decided, apparently, that someone else can worry about it.
Consider a real person again. A trans woman, fifty-eight years old, obtained her Gender Recognition Certificate in 2009. At that point she had lived in her acquired gender for seven years. She paid the fee. She satisfied the panel. She received from the state a formal legal recognition of her identity. She arranged her life — her bank accounts, her passport, her workplace, her relationships, her sense of self — on the basis that the state had recognised her, and would continue to recognise her, as a woman.
In May 2026, she opened the EHRC Code and read that, for the purposes of the central piece of UK equality legislation, the document she had relied on for seventeen years was meaningless. She read that she might be excluded from women’s services on the proportionality test. She read that her exclusion from the men’s service would also be lawful on the same logic. She read paragraph 2.51 and understood that the regulator was not interested in what this meant for the rest of her life. She closed the document.
The Code has nothing to say to her. That, in the end, is what it is most fundamentally about.
Article 8 and the human rights dimension
The Code makes occasional, glancing gestures toward the European Convention on Human Rights. Article 8 is mentioned at paragraph 13.161. Paragraph 13.162 states that asking about sex must be objectively justified to comply with Article 8. There are scattered references throughout to the Human Rights Act 1998. These references are not engagement. They are decoration.
The Code does not address Goodwin v United Kingdom (2002), the foundational Strasbourg authority on the recognition of acquired gender, in any substantive way. The case that produced the Gender Recognition Act 2004 — and on whose continued vitality the credibility of the United Kingdom’s compliance with Article 8 depends — does not appear in the Code’s analytical reasoning. The line of subsequent ECHR jurisprudence on the privacy and identity interests of trans people is not addressed. The Code proceeds as if the Convention dimension were a matter of housekeeping rather than substance.
The Strasbourg Court has held repeatedly that the recognition of acquired gender falls within the scope of Article 8. It has not yet been asked to consider whether a state may grant legal recognition of acquired gender for general purposes under one statute and deny that recognition for the purposes of another. It will be asked, in the years following For Women Scotland and the Code that implements it. When it is asked, the United Kingdom will be required to defend the position that the EHRC has produced.
The Convention defence available to the United Kingdom in such a case rests on the margin of appreciation, on the legitimate aim of protecting women’s single-sex spaces, and on the proportionality of the means chosen. The Code, as drafted, materially weakens each of these defences. It weakens the margin of appreciation argument because it has gone substantially further than the Court’s narrow ruling required. It weakens the proportionality argument because the means it chooses — front-line delegation to lay providers — is plainly less precise and more intrusive than alternatives that could have been adopted. And it weakens the legitimate aim argument because, in its operation across both sides of the trans population, it produces the exclusion of people from services they have a statutory right to use, with no equivalent service available to them.
A Code that took Article 8 seriously would have engaged with these problems. It would have addressed the proportionality of asking about sex in everyday service settings. It would have set out a verification procedure that minimised the privacy intrusion. It would have considered the data protection consequences of recording the resulting information. It would have addressed the specific Convention exposure created by the position of the GRC holder. This Code addresses none of these matters in any substantive way. A claimant in a future Article 8 case will find very little in the Code that protects them. The United Kingdom will find very little in the Code that protects the United Kingdom.
The litigation to come
A Code that protects no one and satisfies no one will be litigated extensively. This is not a prediction. It is an arithmetical certainty.
Some of the litigation is already foreseeable in outline. Service providers who adopt blanket exclusionary policies will be sued by trans claimants for direct or indirect gender reassignment discrimination. Service providers who adopt inclusive policies will be sued by other service users for indirect sex discrimination. Service providers attempting to apply the proportionality test on a case-by-case basis will be sued from both directions, often by claimants who hand the same Code to the same court in support of opposite conclusions. Refuges, in particular, will find themselves on the receiving end of letters before action from claimants of every description.
The Code itself is, in addition, vulnerable to challenge. A Code of Practice issued under section 14 of the Equality Act 2006 may be amended or withdrawn by the Secretary of State, and may also be the subject of judicial review where its content is alleged to be unlawful. Grounds available to a challenger include: incompatibility with Article 8 of the ECHR; misdirection in the application of the Supreme Court’s reasoning in For Women Scotland; failure to take account of relevant considerations under section 149 of the Equality Act 2010, the public sector equality duty; and procedural deficiencies in the consultation process by which the Code was developed. Whether any such challenge succeeds is impossible to predict in advance. That such challenges will be brought is not seriously in doubt.
Beyond the inevitable individual claims and the foreseeable strategic challenges, there is the question of what happens at appellate level. The Code will be cited in the County Court, the Employment Tribunal, the Upper Tribunal, the Court of Appeal and, in time, the Supreme Court itself. Sooner or later — and it will be sooner rather than later — an appellate court will be asked to consider whether the Code’s framework, in whole or in part, is consistent with the Equality Act 2010 and the Human Rights Act 1998. The Code is not legislation; it is guidance. The courts will treat it as such. The likelihood that all of it survives appellate scrutiny intact is not high. The likelihood that significant parts of it are read down, distinguished, or quietly ignored by tribunals reaching for sensible outcomes in particular cases is, on any sober assessment, considerable.
The Code is at the start of its life. It will not reach the end of its life in the form in which it was published.
Conclusion: the day British equality law became unmanageable
The Code’s defenders will say that the EHRC was obliged to take the For Women Scotland judgment into account, that the difficulties in its framework are inherent in the underlying legal position, and that the criticism in this commentary is criticism of the Supreme Court dressed up as criticism of the regulator. That defence does not work.
The EHRC was not obliged to produce a 342-page Code. It was not obliged to extend the For Women Scotland reasoning into every domain of public life. It was not obliged to deputise the proportionality test to every shopkeeper in the country. It was not obliged to publish worked examples that, if applied as written, exclude survivors of domestic violence from the only support service available to them. It was not obliged to draft a verification framework that hands every operator in the country a permission slip to interrogate the apparent sex of anyone whose appearance does not conform to the operator’s idea of what a member of the eligible sex looks like. These were choices. They were the choices of an institution that, faced with a politically charged question, opted to maximise the regulatory scope of its response rather than minimise it; opted to translate a narrow Supreme Court ruling into a sprawling regulatory programme rather than a modest technical revision; and opted to do all of this without the depth of consultation, the precision of drafting or the courage of moral clarity that the seriousness of the subject required.
The choices will be tested. They will be tested by claimants who find themselves at the wrong end of a worked example. They will be tested by service providers caught between obligations they cannot simultaneously discharge. They will be tested by judicial review, by appeal, by ECHR application, and by the slow corrosive work of front-line non-compliance by people whose practical good sense exceeds the Code’s own. They will be tested over years, during which the lived experience of trans people, of women using single-sex services, and of ordinary members of the public moving through public space will all be made worse than they needed to be, in the name of an equality regulator whose Code has, on any honest reading, failed to do its job.
The Supreme Court delivered a narrow ruling on the meaning of a word. The Equality and Human Rights Commission, in its name, has delivered a regulatory framework that no honest reading of the judgment required, that no group affected by it welcomes without qualification, that no service provider asked to apply it can apply with confidence, and that no court will leave standing in its present form. We may yet look back on 6 May 2026 as the day British equality law became unmanageable. It will, in any event, be a day on which the Equality and Human Rights Commission failed the people it exists to serve.
The full text of the Code is available at gov.uk/official-documents. The Supreme Court’s judgment in For Women Scotland Ltd v The Scottish Ministers [2025] UKSC 16 is published on the Supreme Court’s website. References in this commentary to paragraph numbers are to the paragraphs of the EHRC Code as published on 6 May 2026.