Is gender segregation in education discriminatory?


Rachel Barrett and Chris Milsom discuss the recent case of Interim Executive Board of X School v Chief Inspector of Education, Children’s Services and Skills [2016] EWHC 2813 (Admin) in this blog. A copy of the judgment is available here.

Many of us might instinctively think that the answer to the opening question would be ‘yes’. After all, in 2013 when Universities UK sparked controversy by suggesting that gender segregated seating may be permissible at university events,[1] updated guidance from the EHRC was modified was to provide that: “Segregation by gender is clearly not permitted in universities’ normal academic, teaching and research activities … Gender segregation is not permitted in any academic meetings or at events, lectures or meetings provided for students, or at events attended by members of the public or employees of the university or the students’ union”.[2]

But what about single sex grammar and independent schools? There is a specific exemption in the Equality Act 2010 (‘EqA’) that allows single sex schools to have sex or gender specific admissions policies.[3] Any other school which restricts intake to one gender would be acting unlawfully in breach of s.85(1) EqA which prohibits discrimination in school admissions on grounds of sex.[4]

However, in the recent decision of Interim Executive Board of X School v Chief Inspector of Education, Children’s Services and Skills [2016] EWHC 2813 (Admin), the High Court held that an Ofsted report was wrong in stating that gender segregation in a mixed sex school (‘the School’) was discriminatory. The segregation in that case had taken place within the educational setting rather than at the admissions stage. Jay J considered that although the segregation was capable of amounting to a detriment, and it was certainly done because of the students’ sex, it was not discriminatory because nobody had been treated less favourably than anybody else.

The factual background

The School is a voluntary aided faith school for boys and girls aged 4 to 16. Pursuant to its Islamic ethos, pupils are segregated by gender[5] from Year 5, i.e. between the ages of 9 and 16. In effect, it operates as two single sex schools on one site with all lessons, breaks, school clubs and trips facilitated separately. Evidence referred to in the judgment suggests that boys and girls are even prohibited from talking to each other. The education provided to each group is of equal quality.  This is not necessarily unusual. Jay J took judicial notice of the fact that some Orthodox Jewish schools are run along the same lines, and also conducted “brief internet research” (Google?) which showed a number of Christian faith schools have similar practices.

In June 2016, Ofsted carried out an inspection of the School and sent a draft report giving it an “inadequate” rating for a number of reasons. One of the reasons was that the practice of gender segregation was considered to amount to direct sex discrimination in breach of the EqA. The School obtained an interim injunction prohibiting Ofsted from publishing the report, and applied for judicial review on various grounds. The focus of the judgment (and of this blog) was on the School’s contention that the Ofsted inspector’s approach to the EqA was wrong in law because segregation, without more, is not discriminatory.

The legal framework

Schools are subject to the education provisions in Part 6 EqA, and in particular to s.85 EqA:

85 Pupils: admission and treatment, etc.

(1) The responsible body of a school to which this section applies must not discriminate against a person—

(a) in the arrangements it makes for deciding who is offered admission as a pupil;

(b) as to the terms on which it offers to admit the person as a pupil;

(c) by not admitting the person as a pupil.

(2) The responsible body of such a school must not discriminate against a pupil—

(a) in the way it provides education for the pupil;

(b) in the way it affords the pupil access to a benefit, facility or service;

(c) by not providing education for the pupil;

(d) by not affording the pupil access to a benefit, facility or service;

(e) by excluding the pupil from the school;

(f) by subjecting the pupil to any other detriment.

(Emphasis added)

Ofsted contended that the School’s practice of gender segregation denied pupils access to a ‘benefit’ or ‘facility’ (i.e. that of mixing with the opposite sex) and amounted to a ‘detriment’.

The type of discrimination alleged was direct sex discrimination. Under s.13 EqA, direct discrimination occurs where a person is treated less favourably than (actual or hypothetical) others, in materially similar circumstances, would be treated because of a protected characteristic (e.g. sex).

There are some limited circumstances where segregation is expressly permitted under the EqA. For example, sex discrimination in competitive sport (including for children) is allowed where “the physical strength, stamina or physique of average persons of one sex would put them at a disadvantage compared to average persons of the other sex” (s.195 EqA).  Further, as set out above, single sex schools and schools of a religious character are permitted to have gender specific or religious criteria in admissions policies. However, there is no exemption that specifically permits gender segregation in educational institutions for religious reasons.

Schools are also subject to the public sector equality duty (‘PSED’) at s.149 EqA, which obliges them to have due regard, in the exercise of their functions, to the need to: (a) eliminate discrimination; (b) advance equality of opportunity and; (c) foster good relations between persons who share a relevant protected characteristic and persons who do not share it.

The decision

Children at the School are segregated according to sex, thus because of a protected characteristic. Jay J considered that at least in principle the denial of the choice to interact with the opposite sex is capable of amounting to a denial of a benefit or facility for the purposes of s.85(2)(b) EqA, and – which he though a better fit – a detriment for the purposes of subparagraph (f). It is an opportunity “which reasonable people would value”, and there was some, albeit anecdotal, evidence that pupils in the School felt disadvantaged, for example saying that they were not being provided with social skills. The key questions were therefore “whether the denial of this opportunity to both sexes amounts to “less favourable treatment” for the purposes of section 13(1)” or “is one sex being treated less favourably than the other?”.

The judge considered whether there was less favourable treatment because boys lost the opportunity to socialise with girls, and girls with boys which was important in both cases for learning to interact confidently in preparation for their work, education and personal lives after leaving school. However, he rejected the proposition that there was mirrored discrimination, or one treatment resulting in two discriminatory consequences. The judge held that “it is artificial to say that the denial to the boys of the opportunity to mix with the girls (which the latter enjoy as between themselves) is somehow different from the opportunity being denied to the girls. It would only be different if there were some qualitative distinction for these purposes between male and female interaction (each looked at inter se), but in my judgment there is not.” Further, “to the extent that female interaction inter se might be said to be different to comparable male interaction, I would hold that the denial of the one is not less favourable than the denial of the other.” On this analysis, both groups had been denied an equal benefit or exposed to an equal detriment thus the treatment was not less favourable and hence non-discriminatory.

Jay J went on to consider whether segregation was ‘less favourable’ in relation to the girls as part of a structurally disadvantaged social group. He accepted as a general but not a universal proposition that women have been and still are the group with minority power in society. However, he concluded that there was no evidence adduced in the litigation to show that segregation in a mixed school has a greater impact on female pupils, nor to explain how and why segregation would particularly disadvantage girls.

Counsel for the Defendant cited the seam of US case law from Brown v Board of Education of Topeka 349 US 294 (1954), in which SCOTUS overturned its earlier ruling in Plessy v Ferguson 163 US 537 (1896) by holding that “separate but equal” schools for different races were unconstitutional. She submitted that making separate equal provisions for boys and girls, as for different races, cannot be divorced from the historic and current societal treatment of the less powerful group. Jay J found this to be “a powerful submission that cannot be lightly dismissed”. However, he ultimately came to the conclusion that the situations were not analogous. Faith schools do not operate under a compulsory government mandated policy of segregation; the practice is endorsed by the parents and does not reflect the mores and attitudes of wider society. Further, the judge did not accept that segregation resulted in any feeling of inferiority on the part of girls, as was suffered by the black school students in Brown. Lastly, there was no evidence that Islamic or other faith schools viewed girls as inferior or thought that they should be separately prepared for a lesser role in society.

Jay J also concluded that there was no breach of the PSED, which (as far as the June 2016 report went) was advanced on the same basis as direct sex discrimination. There was no evidence that the School has failed to promote equality of opportunity for its female pupils.


This is a finely balanced case capable of more than one reasonable interpretation. Permission has already been given by the Divisional Court for appeal to the Court of Appeal. We consider that there are two respects in which a different conclusion may be reached, by (a) considering the differential impact of separate but ostensibly equal provision; and (b) looking at whether sex is a legitimate ground for difference in treatment, whether or not a more favourably treated comparator can be identified.

  1. (a)   Differential impact

It is trite that treating two parties or groups badly in the same way does not infringe the prohibition on direct discrimination (sometimes colloquially referred to as the ‘b****** defence’ i.e. “I would have bullied a man just as I bullied this woman”). However, it does not follow that where, as here, you have mirrored treatment, the treatment of each group cancels out the other. Girls are denied the opportunity to associate with boys, whereas boys are not. Boys are denied the opportunity to associate with girls, whereas girls are not.

The judge concluded on the evidence before him that there was no difference in the effect of segregation on boys and girls. He had “little doubt” that educational experts would have much to say on the topic, but he had not heard it “within the four corners of this litigation”. However, he did take judicial notice of the fact that women continue to be the disadvantaged or less powerful social group. It takes only inferential reasoning from that position to deduce that segregation is likely to have different effects on girls and boys.

This is, indeed, borne out by research conducted by educational experts. Studies in the USA, Israel and Spain have shown that both boys and girls do better when there are more girls in the classroom; so while girls tend to benefit academically from single sex classrooms, boys do less well.[6] On this basis, the segregated boys are treated less favourably because statistically they can be expected to do less well academically as a result of being separated from their female peers.

The treatment has different implications for girls, but may still be ‘less favourable’ in non-academic respects. Girls belong to the sex that is (as Jay J accepted) still less advantaged in UK society; depending on their future career paths they are likely to face a gender pay gap and lack of female role models in senior positions in comparison to men. It is therefore important that they develop confidence in mixed sex environments during their formative years, for different and gendered reasons to that of the boys.

If as a result of segregation the boys are treated less favourably than girls in some respects more favourably in others, and vice versa for girls, they are both discriminated against. A court will not weigh up the advantages and disadvantages to decide whether overall they come out equal, or one or the other on top. For example, in Ministry of Defence v Jeremiah [1980] QB 87, men in a factory were treated less favourably than the women in that they performed additional work in dirty conditions, but were treated more favourably in that they had the opportunity to do overtime. Discrimination was made out; one aspect did not cancel out the other.

  1. (b)   Sex as a ground for different treatment

Jay J’s emphasis on each group as an exact comparator does not reflect that men and women are perceived and stereotyped differently, and that difference in treatment based on sex should be stringently scrutinised. He accepted that a hypothetical case of segregating Muslims and Hindu children but otherwise apparently treating them equally “would be an egregious case of religious discrimination” because the inference would be that the more powerful group was imposing its will on the weaker with correlative express or implied disadvantages. That does not fully explain why as a society we see segregation as inherently undesirable as between social groups, absent specific reasons such as physical strength in competitive sport. Even if, in this hypothetical religiously divided school, neither group was exploiting superior power to impose its will on the weaker group, distinction based on an irrelevant (to school education) characteristic is antithetical to this our non-discriminatory principles.

In her opinion in Bartsch v Bosch und Siemens Hausgeräte (BSH) Alteredsfürsorge GmbH C-427/06 [2009] 1 CMLR 5 (a case concerning age discrimination in pension provision), AG Sharpston considered the evolving nature of the concept of equality:

“A classic formulation of the principle of equality, such as Aristotle’s “treat like cases alike” leaves open the crucial question of which aspects should be considered relevant to equal treatment and which should not. Any set of human beings will resemble each other in some respects and differ from each other in others. A maxim like Aristotle’s therefore remains an empty rule until it is established what differences are relevant for the purposes at hand. For example, if we criticise a law banning redheads from restaurants as being unjust, that is based on the premise that, as regards the enjoyment of a meal in a restaurant, hair colour is irrelevant. It is therefore clear that the criteria of relevant resemblances and differences vary with the fundamental moral outlook of a given person or society.”

Is gender a relevant difference on which separation in the classroom environment may legitimately be based? Certainly race is not, for the reasons advanced in Brown and subsequent case law. Neither would sexual orientation be an acceptable dividing line, even though gay and straight pupils would both be deprived of each other’s companionship. Severe disability may be a relevant difference, where suitable provision cannot be made in a mainstream school.[7] Arguably gender may be, for example to benefit girls’ academic attainment. However, this cannot be taken for granted and should not escape scrutiny merely because both genders are being deprived of each other’s company.[8]

Parliament has made distinct provision in the EqA to permit single sex schools. If parents and educators wish to make provision for single-sex education, they can do so by creating two separate schools and, thereby, benefiting from this exemption. However, there is no such exception for gendered segregation (in both teaching and social interaction) in a mixed sex environment. Direct discrimination under s.13 EqA cannot be ‘justified’ – for example, for religious reasons – as lawful.

Notwithstanding our points above on the different impacts of segregation on girls and boys, the emphasis of this type of analysis is not on identifying the comparator group who are treated more favourably, but rather asking whether the characteristic in question (here, sex or gender) is a legitimate basis for difference in treatment. As submitted by Counsel for the Defendant in the case (but not accepted by the judge), “the court should not get too hung up on the question of a comparator”.

In the EHRC guidance for universities ‘Gender Segregation at Events and Meetings’ referred to above, a comparator suffering mirrored treatment is considered to suffice to show direct discrimination:

Any seating arrangement amounts to direct discrimination if it results in disadvantage to any participant (actual or potential) because of gender. Disadvantage occurs wherever an individual, male or female, might reasonably take the view that they had been treated less favourably because of gender. Thus a woman who is not permitted to occupy a particular area of the lecture hall because it is reserved for men is disadvantaged because she cannot sit wherever she chooses (or vice versa). The provision of a non-segregated area within the lecture hall does not prevent that disadvantage arising to either gender.”

This is of course not binding legal authority, but strikes the authors of this article as a helpful analysis of gender segregation in education.

Lastly, we query whether the School was in breach of its PSED duty, not just to promote equality of opportunity (as advanced in the June 2016 report and rejected by the judge) but also to foster good relations between different social groups. Intergroup contact reduces prejudice.[9] Men and women who grow up with members of the opposite sex as classmates and friends rather than a distantly observed separate species (whether from Mars or Venus) will find it more natural to view and treat each other as equal fellow humans.

Rachel Barrett and Chris Milsom are barristers who specialise in discrimination law. Please contact the clerks on 0207 827 4000 if would like more information about working with them or any other barristers at Cloisters. You can find out more about Cloisters at cloister.code.





[3] Para 1 of sch.11 EqA. Local authorities are also exempt from the prohibition on sex discrimination in the exercise of the authority’s functions in relation to the establishment of a school – para 8 of sch.3 EqA.

[4] Also race, disability, sexual orientation, gender reassignment, pregnancy and maternity – EqA ss.4 & 84; and only schools with a religious character have an exemption to discriminate on grounds of religion or belief – s.89(12) & sch.11 EqA.

[5] Gender for the purposes of the school’s policy being binary male or female. The EqA refers to the protected characteristic of ‘sex’, but the concept encompasses both gender and biological sex.

[6] Iris Bohnet What Works: Gender Equality by Design (Harvard University Press, 2016) p.224; the studies can be found at,, and Lavy & Schlosser, “Mechanisms and Impacts of Gender Peer Effects at School,” American Economic Journal: Applied Economics 3 (2011): 1-33.

[7] More favourable treatment of disabled persons is permitted by s.13(3) EqA.

[8] Neither, we suggest, can parental consent be a relevant factor: consent does not vitiate discrimination (see Ministry of Defence v Jeremiah), and certainly not consent on another’s behalf.