Identifying Coercive Control in the Courts

The BBC reported today that new research from Manchester Metropolitan University shows a strong link between coercive control and murder. This echoes the research done by Professor Jane Monckton-Smith in her book In Control: Dangerous Relationships and How They End in Murder. 

It is probably no surprise, at least to feminists who follow such research closely, that abusers sometimes escalate to femicide. 

The question is what can be done about it – and why legal remedies are struggling to improve the figure of roughly 100 women a year murdered by an abusive partner or ex partner. 

Both the criminal and family courts have found it difficult to draw a bright line between behaviour that is unpleasant or intolerable and behaviour that is coercive and controlling. 

As  Peter Jackson LJ commented in Re L (Relocation: Second Appeal) [2017] EWCA Civ 2121 at paragraph 61:

“Few relationships lack instances of bad behaviour on the part of one or both parties at some time and it is a rare family case that does not contain complaints by one party against the other, and often complaints are made by both. Yet not all such behaviour will amount to ‘domestic abuse’, where ‘coercive behaviour’ is defined as behaviour that is ‘used to harm, punish, or frighten the victim…’ and ‘controlling behaviour’ as behaviour ‘designed to make a person subordinate…’ 

This was the approach echoed by Hayden J in F v M [2021] EWFC 4 at para 4:

“…In the Family Court, that expression [‘coercive and controlling behaviour’] is given no legal definition. In my judgement, it requires none. The term is unambiguous and needs no embellishment. Understanding the scope and ambit of the behaviour however, requires a recognition that ‘coercion’ will usually involve a pattern of acts encompassing, for example, assault, intimidation, humiliation and threats. ‘Controlling behaviour’ really involves a range of acts designed to render an individual subordinate and to corrode their sense of personal autonomy. Key to both behaviours is an appreciation of a ‘pattern’ or ‘a series of acts’, the impact of which must be assessed cumulatively and rarely in isolation.”

In the criminal sphere, a similar description is offered to prosecutors in guidance. 

There is of course a definition of the offence, provided s.76 Serious Crime Act 2015. The CPS guidance explains that a person will be guilty of the offence of coercive control in the following circumstances:

An offence is committed by A if:

  • A repeatedly or continuously engages in behaviour towards another person, B, that is controlling or coercive; and
  • At time of the behaviour, A and B are personally connected; and
  • The behaviour has a serious effect on B; and
  • A knows or ought to know that the behaviour will have a serious effect on B.

A and B are ‘personally connected’ if:

  • they are in an intimate personal relationship; or
  • they live together and are either members of the same family; or
  • they live together and have previously been in an intimate personal relationship with each other.

There are two ways in which it can be proved that A’s behaviour has a ‘serious effect’ on B:

  • If it causes B to fear, on at least two occasions, that violence will be used against them – s.76 (4)(a); or
  • If it causes B serious alarm or distress which has a substantial adverse effect on their day-to-day activities – s.76 (4) (b).

The guidance goes on to say that “For the purposes of this offence, behaviour must be engaged in ‘repeatedly’ or ‘continuously’. Another, separate, element of the offence is that it must have a ‘serious effect’ on someone and one way of proving this is that it causes someone to fear, on at least two occasions, that violence will be used against them. There is no specific requirement in the Act that the activity should be of the same nature. The prosecution should be able to show that there was intent to control or coerce someone.”

In the criminal court, the offence must be proved beyond reasonable doubt. Where the allegation is made in a family court, the pattern of behaviour must be proved on the balance of probabilities. 

It is readily apparent that it will be difficult in very many cases to draw a line between “bad behaviour” and “coercive control.” 

One complicating factor could well be the expectations of the judge or jury as to the victim’s response. Some women will respond to controlling or coercive behaviour by becoming meek, withdrawn, and afraid – the “freeze” response. Others will respond to violence with retaliatory violence or angry outbursts – the “fight” response. Perpetrators may provoke such a reaction and then accuse their victims of being the ones engaging in coercive control, a pattern known as DARVO (Deny, Attack, Reverse Victim and Offender). The courts must be alive to unexpected reactions from victims: just because she has shouted at him to “fuck off” after months of belittlement or financial abuse doesn’t mean that “she gave as good as she got” and that her sense of autonomy must therefore be intact. 

There are no easy answers to how coercive control is identified, but early intervention would save lives. Education on identifying and preventing it, for prospective victims and prospective perpetrators, as well as those living in abusive homes, should begin in PHSE in adolescence rather than waiting for abuse to occur, when the post-mortem on events may tragically not be metaphorical.

After Forstater: elephants and elephant traps 

This is the text of a talk I gave on Wednesday evening to employment law solicitors at my Chambers.

I should start by acknowledging the elephant in the room. I broadly share the belief that lost Maya Forstater her work with CGD: namely that biological sex is real, important, immutable and not to be conflated with gender identity. We’re not going to be discussing the substance of that belief except tangentially, but inevitably there are various respects in which the position I take will affect the way I talk about my understanding of the law. 

As ever: views my own.

Staying with the elephant theme, the question I want to address is: “what are the main elephant traps for your clients in this area, and how do they avoid them?” I’m going to address that partly by reference to five cases that have been fought to a conclusion in the ET or on appeal over the last couple of years. So first, a quick outline of those cases. 

The five cases 

An employment tribunal held in December 2019 that Ms Forstater’s gender critical belief was not protected under s.10 of the EqA because it was “not worthy of respect in a democratic society” (or “WORIADS” as it’s come to be known). In June 2021, the EAT allowed Ms Forstater’s appeal, so that the case could be heard on its merits. In July this year, the ET held that CGD’s decision not to renew her contract had been because of her protected belief, and was therefore unlawful direct discrimination. 

In Mackereth, the EAT upheld a tribunal’s decision that the DWP’s treatment of a medical assessor who refused to use the preferred pronouns of service users was not discriminatory. 

In Bailey v Stonewall Equality Ltd & ors, the barrister Allison Bailey sued her chambers and Stonewall for belief discrimination. She won part of her claim against her chambers and lost other parts. She lost the claim against Stonewall, and is appealing that part of the tribunal’s judgment. 

The other two cases are V v Sheffield Teaching Hospital (which Anya Palmer has written about in more detail here) and Taylor v Jaguar Land Rover. In V, a tribunal upheld the claimant’s complaint that questioning him about his habits in relation to wearing underwear at work was discrimination on grounds of gender reassignment. He had been seen naked from the waist down in a women’s changing room. In Taylor, a tribunal found a number of complaints of harassment and direct discrimination proved by a trans-identifying male employee. 

Between them, these five cases shed quite a bit of light on the elephant traps I want to talk about. 

The elephant traps 

The “social media policy” fallacy – treating some beliefs as more equal than others

All beliefs that pass the 5 Grainger tests needed to qualify for protection are of equal status. Employers are entitled to ask their employees not to proselytise at work; and they will often be entitled to place some restrictions on their employees’ public statements outside the workplace. Exactly how far those restrictions can go will depend on a range of factors – the ease with which the employee can be identified as such, her seniority; the nature of her role, and so on. Judges and civil servants can be required to keep pretty silent, in public, on matters of political controversy; supermarket checkout staff not so much. 

There has been quite a lot of comment on Forstater to the effect that CGD’s difficulties could have avoided if only they had had a robust social media policy in place. 

There are two problems with that. The first is that Forsater’s engagement, though direct, was pretty measured.  That means that any social media policy sufficiently restrictive to silence her on the subject of GRA reform would have had to be draconian across the board. If it had singled out “gender critical” engagement for prohibition, that would have been discrimination just the same. You can’t make discrimination disappear by making it your policy to discriminate, and then saying you acted as you did not on the prohibited ground but in obedience to your policy. An employer could in theory decide on the draconian route, and just purport to put all political or contentious social media engagement out of bounds. But trying to enforce such a draconian policy would be likely to have a high cost in both management time and industrial relations. And an employer that dismissed for breach of such a rule might well find it hard to defend as consistent with the employee’s article 10 right to freedom of expression, which will be part of “all the circumstances” a tribunal has to consider when ruling on the fairness of a dismissal. 

On the other hand, if an employer writes a draconian policy but only enforces it reactively when staff members or third parties take strong exception to the expression of particular views – well, the problem with that should be obvious. Effectively you’d be letting the mob decide which opinions may be expressed. If the mob discriminates, you discriminate. 

A related elephant trap may be concealed in a more limited neutral-looking policy. Suppose your policy says something like this:  

You must not make any social media communication that could damage our business interests. 

It looks even-handed and fairly light-touch. But if what this formula really means is that employees mustn’t express unfashionable views because third parties might object, that won’t preserve the employer from a finding of discrimination. A more familiar parallel may help make this vivid: an airline can’t get away with saying,  “Well of course we know that women make perfectly good pilots, but we don’t employ women to fly planes because our passengers wouldn’t feel safe and would vote with their feet.”

So what should social media policies look like? I’d suggest that a sensible policy for most organisations will simply ask senior and middle-ranking employees to make it clear that the views they express are their own, and to express themselves lawfully and reasonably courteously in any event. There’s nothing conceptually difficult about that, though an organisation faced with a social media pile-on may be called upon to hold its nerve.

That takes me to my next elephant trap. 

Running scared 

This was illustrated by CGD’s conduct after Maya Forstater started to engage in the debate about the proposal to reform the GRA to bring in self-ID. 

It’s notable from the evidence quoted in the judgment that some managers were initially nonplussed: they weren’t sure what all the fuss was about. One even admitted at an early stage that he wasn’t sure whether or not he agreed with Forstater. But as the  campaign against her intensified, and they became aware of internal reactions to her tweeting described by one witness as “visceral”, they fell into line. 

Their problem seems to have been a disinclination on the part of managers to look behind claims to be offended by Ms Forstater’s tweets, and make up their own minds whether she had said anything genuinely unacceptable. The high-water mark of the evidence against her was that she had described a man known sometimes as Philip Bunce and sometimes as Pips Bunce as a part-time cross-dresser in the context of a discussion about whether he should have accepted an award for women in business. 

“Cross-dresser” is a term that can be found in many LGBTQ+ organisations’ glossaries, and Bunce is a man who sometimes but not always cross-dresses at work – so for my own part I find it difficult to see what was wrong with Forstater’s  description. But that’s not to say it was completely fanciful to think her phrase a bit rude: the tribunal itself was split on that question, EJ Glennie and Mr Miller taking the view that it was uncomplimentary and dismissive but not in all the circumstances inappropriate or objectionable; the third member, Ms Carpenter did think it objectionable although that didn’t affect Ms Carpenter’s view of the result in the case.  

Sensible managers would have given the complaints short shrift. Employees are not entitled to demand that their employers protect them from having to work with people they disagree with –  or even with people who are sometimes a bit rude about third parties on social media.

That leads to my next elephant trap. 

“Bring your whole self to work” 

It’s become fashionable for HR policies to talk about making everyone feel pyschologically safe and able to bring their whole selves to work.

This may be some of the worst advice ever given to employees. 

None of us should bring our whole selves to work.  It’s perfectly fine to be an enthusiastic amateur opera singer, ju-jitsu practitioner or free-climber on your own time – but if you sing opera, wrestle your manager or literally climb the walls at the staff meeting, it won’t go well. Mr Pay, the Claimant in Pay v Lancashire Probation Service was dismissed because of what he did on his own time. If he had brought his whole self to work, it wouldn’t have taken the full intellectual heft of the Court of Appeal to spot that dismissal was fair.

More seriously: this kind of messaging is calculated to lead employees to expect to be allowed to police their colleagues’ beliefs and opinions. That’s not going to work in a diverse society. Maya Forstater’s belief that the differences between male and female bodies sometimes matter seem to have been bitterly offensive to some of her colleagues – but no doubt the opposing belief that men can be lesbians, and women and girls are not entitled to any reliable privacy from men was bitterly offensive to her. You can play the same game with many irreconcilable beliefs. An ethical vegetarian may think I am little better than a murderer because I eat meat; a religious colleague may think I am destined for hell because I don’t believe in God, an environmentalist that I’m a vandal because I drive a car. 

That’s all ok – or it should be. These kinds of incompatibilities of belief may sometimes make friendship difficult, but they shouldn’t impede working together, provided everyone respects everyone else’s freedom of belief, conscience and speech. 

So the message for employers is: don’t write policies that give employees the impression that they can expect their colleagues to share their beliefs, or even pay lip-service to them. But do make sure that employees don’t proselytise or try to impose their own beliefs on others in the workplace. 

Of course, elephant traps are not just there for employers. The next one is for employees:

Being a martyr

Dr Mackereth was employed by the DWP to make medical assessments for the purposes of disability-related benefits. He was a Christian whose rejection of genderist beliefs had biblical roots. He made it clear in the course of his induction that he could not in conscience use pronouns for service-users other than those indicated by their sex. 

The DWP sought to explore with Dr Mackereth the parameters of his position, seemingly with a view to retaining his services if it could, but Dr Mackereth resigned – saying that he believed he was being dismissed – while that process was ongoing. The EAT confirmed that the tribunal had permissibly found that the DWP’s conduct had been a response not to his beliefs but to the way in which he had indicated he was determined to manifest them.  

Dr Mackereth seems to have jumped early on to the conclusion that he was bound to be dismissed, and to have been unwilling to engage constructively with the DWP’s attempts to find an accommodation. It’s not obvious that that was necessarily a lost cause: Dr Mackereth was willing to use clients’ preferred names, and in 1:1 meetings you might think it wouldn’t be too difficult to swerve the question of pronouns altogether. 

If you are advising an employee at an early stage of a dispute of this nature, I suggest there are four key lessons from Mackereth: 

1. Stay calm, and assume your employer is acting in good faith until the evidence to the contrary is overwhelming.

2. Don’t force matters to a head: be open to pragmatic work-arounds that respect others’ conscience and belief as well as your own. 

3. Decide on your own red lines and communicate them clearly; but – crucially – 

4. Don’t jump before you are pushed. If someone is going to decide that your beliefs can’t be accommodated in the workplace, leave it to your employer to make that decision. 

That takes me to my final elephant  trap:

Confusing the right not to suffer GR discrimination with a right to be treated as the opposite sex

The case that illustrates this is V v Sheffield Teaching Hospitals NHS Foundation Trust. It’s first instance ET decision, so it has no value as precedent  – and in any case it is in my view pretty obviously wrong. But it serves as a cautionary tale. 

V, a trans-identifying male, applied for a job as a catering assistant. He was given permission to use the women’s toilets and changing and showering facilities from the start of his employment, and his female colleagues were told that that was what he would be doing, and given bespoke training – before V started work – that seems to have been designed to ensure that they didn’t raise objections. The judgment of the ET is silent as to what if any medical treatment he had undertaken, but it is clear from how matters developed that he was obviously male. 

V resigned a little over a year after he started work, and made a number of complaints against the hospital, including complaints of gender reassignment discrimination. 

He made a number of complaints. The only one that succeeded  arose as follows. In June 2021, there was a report to a manager that V had been seen naked from the waist down in the women’s changing room. On a previous occasion he had remarked to a colleague that he was hot and sweaty and had taken his underwear off, making a wringing motion with his hands. A manager asked him in a meeting about whether he was in the habit of removing his underwear, and the tribunal found that that question was asked because of his gender reassignment, and was to his detriment. He therefore succeeded to that extent in his complaint of discrimination. All his other complaints were dismissed. 

My elephant trap is evident in the behaviour of both the hospital, and the tribunal. 

The tribunal approached his complaint of discrimination on the basis that if he had not been a transsexual, he would not have been asked whether he was in the habit of removing his underwear. The comparator used by the tribunal is what it calls a “cisgender woman” in a similar state of undress. 

If V had had a gender recognition certificate, there would be a respectable argument that a woman would be the correct comparator; though I think the better view is that even so, the correct comparator is a man who is not trans.  But there is no suggestion in the judgment that V had a GRC, and on the assumption that he didn’t, the tribunal certainly chose the wrong comparator. If you want to know whether V was asked the question because of his trans status,  you need to think how the employer would have treated a man who was not trans who was seen naked from the waist down in the women’s changing room. 

I think the same error underpins the hospital’s approach to V’s use of the women’s facilities. It seems to have assumed that to deny him the right to do so would have been discrimination on grounds of his gender reassignment, and unlawful. But it would have been neither. If he had been excluded – like any other man – that would not have been because he was trans, but because he was a man. It would not have been sex discrimination because there were equivalent facilities for men. And it could not have been unlawful indirect discrimination on grounds of gender reassignment because it was obviously justified in the interests of protecting the privacy and dignity of his female colleagues, and complying with the obligation under the Workplace (Health Safety and Welfare) Regulations 1992 to provide separate male and female facilities. 

It’s my view that this is the legally correct answer to the toilet conundrum: the women’s facilities in a workplace are for the use of women only, and trans-identifying males should be permitted to use either the gents’ or a single-occupancy unisex toilet. And that rule should be applied irrespective of whether a trans-identifying male has a GRC or what if any medical treatment or surgery he may have had: his female colleagues are entitled to have their privacy respected. 

But my view isn’t without its vulnerabilities.  In Croft v Royal Mail, the Court of Appeal ruled that a pre-operative transsexual had lawfully been refused use of the women’s toilets, but also suggested that at some stage of his transition he would have to be treated as a woman. Unfortunately the judgment doesn’t then offer employers the slightest assistance in identifying when that stage is reached. That was in 2003. By 2018, an ET in Birmingham felt able to treat it as self-evident that refusing a trans-identifying man use of the ladies’ was unlawful discrimination even in circumstances where – as is spelled out in the judgment – he had had no surgery and had no intention of undergoing any in the future. That case, Taylor v Jaguar, was a first instance judgment, and in my view clearly wrong in this respect – but it was not appealed. 

So what is the right advice for an employer faced with the toilets conundrum? How does it minimise the risks of being sued? 

The only truly safe option is the kind of facilities that we have here in Chambers: single-occupancy toilets which may be badged as male and female, but for which there’s really not much problem if someone uses the toilets for the opposite sex – anyway provided they aim straight, and raise the seat if appropriate. But suppose what you have is separate halls of cubicles separated by flimsy partitions; and one or two single-occupancy accessible toilets? Suppose that for reasons of cost or space or both you can’t remodel them. If you let a trans-identifying male use the ladies’, your female staff may sue you. If you offer him the use of the accessible toilets instead, he may sue. 

I have nothing very comforting to say here. There’s no binding case law. The remarks in Croft are both Delphic and obiter. And whatever you do, someone’s going to be furious with you. 

I think the best I can offer employers faced with a toilets or changing rooms problem is to suggest that should do what they think is right. That sounds trite, but if you’re likely to end up in court whatever you do, you might as well at least take a decision that feel able to defend wholeheartedly.

If an employer is struggling to form an intuition about what doing the right thing looks like, a good start might be to start not by telling its female staff how they ought to feel about sharing intimate spaces with a trans-identifying male, but asking them how they do feel. Given the climate of fear that’s been generated about admitting you don’t think men can literally become women, it’s probably a good idea to do that anonymously.  

Notes, questions and some links.

I am grateful to Andrew Allen KC, who shared this event with me and provided a very helpful summary of the legislative background and the case law. This blog only reflects what I said.

There were questions about pronouns in email signatures, and “misgendering”. For more on those subjects, see

Is “misgendering” always harassment?

More on “misgendering”

Yet more on misgendering

Grammar and grievance

I was asked about my view on taking HR advice from organisations like Stonewall; on that, see generally Submission and Compliance.

And finally, I was asked an interestingly difficult question by Melanie Field of the EHRC about the meaning of “sex” in the Equality Act. I hope to do justice to that in a future blog.

Post-script: that “future blog” is now here.

Prison Allocation: How Is It Done?

On 8 September 2022 the Daily Mail reported that Sally Dixon, convicted of multiple sex offences against children between 1989 and 1996, had been sentenced to two consecutive nine year terms to be served in the women’s prison estate. This information was newsworthy because Sally Dixon was born male, committed those offences as a male and, as no Gender Recognition Certificate has been issued, remains legally as well as biologically male. 

Dixon will reportedly go to HMP Bronzefield, a women’s prison and young offender institution. Bronzefield also has a mother and baby unit, accommodating babies up to 18 months with their mothers. 

The Mail reported that it was the judge who sent Dixon to a women’s prison, which is wrong. A sentencing judge determines the sentence, but has no role in deciding where it will be served.

The rules on how this is managed are set out within the Ministry of Justice’s policy on The Care and Management of Individuals Who Are Transgender

When a transgender prisoner is identified, a Local Case Board is convened. If the issue is relatively simple (e.g. a female prisoner who identifies as male or non binary but has no GRC and wants to stay in the female estate) then the Local Case Board will complete the process. However if it is more complicated then a referral will be made to a Complex Case Board. This includes cases where someone wants to be placed in the estate of the opposite sex, as has happened here. 

The reader may initially be relieved to hear that the policy claims that “Decisions are free from bias, follow a clear, recorded process and are undertaken by staff who have a sound basic awareness of transgender identity.” The footnote mutters that this sound basic awareness is gleaned from an online e-learning module.

So what is the basis on which the decisions are made? The risks presented both to and by the transgender prisoner must be considered, as well as the prisoner’s own views. It is not the case that the prisoner gets a free choice, nor that a judge has magical dispensing powers. 

The policy provides: 

Decisions must be informed by all available evidence and intelligence in order to achieve an outcome that balances risks and promotes the safety of all individuals in custody as set out below19.

Potential risks to the individual from others, or personal vulnerabilities of the individual, related to: (*indicates critical factors)

*Mental health and personality disorder;

*History of self-harm;

*Anatomy, including risk of sexual or violent assault

*Testimony from the individual about a sense of vulnerability, e.g. in a male
environment, in a particular prison, or from a particular prisoner or group of other prisoners;

*Risk of suicide;

*Medication including the absence of medication and the impact of known side effects

*History of being attacked, bullied or victimised;

*Intelligence including evidence of coercion, manipulation, or threats towards the individual

Family circumstances/relationships


Physical health

Learning disabilities or difficulties.

Potential risks presented by the individual to others in custody and an AP related to: (*indicates critical factors)

*Offending history, including index offence, past convictions and intelligence of potential criminal activity- e.g. credible accusations.

*Anatomy, including considerations of physical strength and genitalia;

* Sexual behaviours and relationships within custodial/residential settings;

*Use of medication relating to gender reassignment; and use of medication generally;

*Past behaviour in custody, the community, in the care of the police, or in the care of prisoner escort services;

*Intelligence reports;

*Evidence of threats towards others;

*Mental health and personality disorder;

Learning disabilities or difficulties;

Substance misuse.

Views/characteristics of the individual: (*indicates critical factors)
*Birth, legal and presented gender;

*Strength of confirmation of presented gender, including medical treatments and full
evidence of gender identity (such birth certificate, or a GRC)

*View on establishment allocation, prison management and lifestyle.

4.19  Whilst the view of the individual on location should always be taken into account, this view must be put into the context of any risks that may be posed to the individual by others (including the risk that they could be threatened or manipulated into giving that view) and the risk that could be posed by the individual to others, whether in the men’s or the women’s estate.

This was the policy that was  unsuccessfully challenged in R (FDJ) v SSJ [2021] EWHC 1746 (Admin). The court was careful to set out its parameters: it was assessing the lawfulness, not the desirability, of the policy [para 72]. On the subject of the Case Boards, the court held: 

“The LCBs and CCBs are expert multi-disciplinary panels. Their members are the persons best placed to assess the risks, and determine the appropriate management of those risks, in a particular case. Those members will surely be well aware of the vulnerabilities of the women who are held in the female prison estate, and of the fear and anxiety which some of them will suffer if a transgender woman, particularly one with male genitalia and/or with a history of sexual or violent offending against women, is accommodated in the same prison. The members are expressly required by the Care and Management Policy[2] to take into account – amongst other relevant factors – the offending history of the transgender woman concerned; the “anatomy, including considerations of physical strength and genitalia” of that person; and the sexual behaviours and relationships of that person. They can in my view be expected to be astute to detect any case of a male prisoner who, for sinister reasons, is merely pretending to wish to live in the female gender.”

In other words, the LCB and CCB must consider the offending history of the prisoner in question when coming to their decision. They are not allowed to ignore it. 

One can perhaps see how a panel might reach the view it did in Dixon’s case. Transition here had begun in 2004 – after the offending period but many years before prosecution –  which suggests the transition was not a cynical attempt to game the system. The panel is entitled to take into account the nature of the offence, but would have to set against that that the last known offending was 1996. Intelligence as to ongoing offending can be taken into account. Sexual offending is notoriously hard to rehabilitate, but it seems unlikely that a panel would infer that there was a present risk simply from the nature of the offending if the last known or even suspected offending was quarter of a century earlier. Add to that the fact that the “risks to” Dixon  included a history of being attacked, bullied or victimised: former friends were reported as having embarked on a course of harassment for which they were sentenced in 2016

All of this leads to a wholly unsatisfactory situation in which a person biologically and legally male, with a history of repeated sexual offending, imprisoned for sexual offending, is nevertheless assessed as presenting an acceptable level of risk to female prisoners. Dixon may not present a significant risk to the women in prison in terms of future offending. But there remains the fact that members of the female prison population are  disproportionately likely to have suffered childhood sexual abuse and other forms of male violence. At best, requiring them to share accommodation with a male who has committed Dixon’s crimes is unlikely to do anything to assist their recovery or  rehabilitation, or with restoring their fragile trust in a system which, too often, has already failed them again and again.

As was highlighted in FDJ [para 82] the policy does not require the LCB or CCB to take into account the “the vulnerability of women prisoners, their frequent experiences of sexual assaults and domestic violence, and the fear and anxiety they may experience as a result of sharing accommodation and facilities with transgender women.”  Lawful though the policy is, it demonstrates an asymmetry which underpins the assumption that the integration of transwomen with women in the prisons’ estate is a desirable objective. Certainly this was the starting point of the court in FDJ. It is, in our view, profoundly sexist but, for good or ill, not all sexism is actionable.

Obviously women prisoners do not have to be assessed in the same way. But if a similar checklist were applied to them, the panels would be reminded of the practical context of their decisions, and be less susceptible to the inequity of awarding primacy to the psychological comfort of male born prisoners over the psychological, physical and sexual safety of women. 

Perhaps it is time that the policy began to require consideration of the vulnerability of women prisoners.