Sex segregation and equality in a multicultural society : inferiority as a standard for legal acceptability

Theory, Achter Sint Pieter 200, 3512 HT Utrecht, tel.: (31) (30) 2537106, email: M.vandenBrink@uu.nl Prof. Dr Titia Loenen, Professor of Legal Theory/Gender & Law, Utrecht University School of Law, Institute of Legal Theory, Achter Sint Pieter 200, 3512 HT Utrecht, tel.: (31) (30) 2537053, email: M.L.P.Loenen@uu.nl Dr Jet Tigchelaar, Assistant Professor of the Philosophy of Law, Utrecht University School of Law, Institute of Legal Theory, Achter Sint Pieter 200, 3512 HT Utrecht, tel.: (31) (30) 2537191, email: H.Tigchelaar@uu.nl


Introduction
Sex-segregated facilities have largely disappeared in the Netherlands and we have the impression that similar developments are under way in other Western countries.Whatever the case may be elsewhere, developments with which we are familiar in the Netherlands provide the point of departure for this contribution, if not exclusively.In the Netherlands, most surviving sexsegregated facilities and practices have to do with morals (bathrooms) and/or with assumed biological differences between the sexes, such as physical strength (sports).They go largely unchallenged.
Interestingly, the downward trend of sex-segregated facilities (e.g. the abolition of singlesex schools, a significant decrease in single-sex hospital wards and all kinds of single-sex private associations, such as the scouting movement) seems to have ended and even to be on the reverse.The renewed demand for sex-segregated facilities is partly caused by religious and cultural beliefs and practices brought along by immigrant groups, such as the wish to avoid physical contact, for instance shaking hands, with persons of 'the opposite sex'.Simultaneously, however, there seems to be a revival of identity politics among women, leading to demands for separate facilities, such as an all-female editorial board of a feminist monthly magazine, or taxis for women driven by women (the so-called lady cabs).A third category of demands arises from a desire felt by some people to redress the effects of neglected 'natural' differences between men and women in the equality discourse.Boys are not girls, so the argument goes, and therefore, for instance, they suffer from the feminisation of the teaching profession; if we want to raise real boys we will have to ensure their own specifically geared educational methods including male teachers et cetera.
In public debate, demands for sex-segregated facilities sometimes meet with cold rejection (separate integration courses for male and female migrants in the Netherlands), whereas at other times, they merely seem to give rise to mild interest (lady cabs) or sympathy (maths classes for girls only).
How should we assess these old and new forms of sex segregation?Which demands for sex-segregated practices are acceptable?According to which standards and on which conditions, and without prejudice against unfamiliar cultural or religious perspectives?
In this contribution we start from the premise that mixed sex facilities and practices are and should remain the point of departure in contemporary Western society.Given the crucial role of the notion and practice of 'separate spheres' for the male and female sex in the long history of women's subordination by men, sex segregation has to be viewed with suspicion and can only be allowed as an exception.By implication, segregated practices are in need of justification, not the other way round.
From a legal point of view, any justification for state-supported, sex-segregated practices will have to meet equality and non-discrimination standards as guaranteed in, for instance, international and regional human rights conventions and/or national constitutions. 1 But how should we decide whether sex segregation indeed infringes these standards?Which criteria or tests should be applied?
In this paper we will formulate such a test and we will assess its implications and merits by applying it to several topical cases of sex segregation.The test draws on anti-domination or anti-subordination theories as developed in, among others, feminist and critical legal theory as well as on the famous landmark cases of the US Supreme Court regarding segregation on the basis of race and sex.To decide whether or not race or sex segregation is acceptable if measured against general legal equality and non-discrimination clauses,2 both sources seem to focus on a single central standard: does segregation create or perpetuate the legal, social, and/or economic inferiority of racial minorities or women respectively?At the end of the day, generally speaking, this indeed seems to be what equality and non-discrimination standards are all about.
After exploring the main features of this standard and formulating what we call 'the inferiority test' (Section 2), we will apply this test to a number of topical cases of sex segregation.We have selected three types of cases: 1.
a traditional case of rather uncontested sex segregation: segregated sports, more specifically amateur football; 2.
a more controversial case that seems to be on the rise once again in several countries: sexsegregated education; and 3.
the highly controversial case of sex-segregated integration courses and other welfare services in the Netherlands.
By choosing both contested and less contested cases we hope to avoid applying double standards to familiar and unfamiliar forms of sex segregation.For each case we will discuss what the application of the inferiority test leads to; would the sex-segregated practices be acceptable or not?
In addition we will pay specific attention to the merits of the test in a multicultural context: is the inferiority test able to deal with multicultural complexity?As many of the new claims for sex segregation may involve not just gender hierarchy but also ethnic, cultural and/or religious hierarchies, we will critically assess whether the application of the inferiority test to practices of sex segregation ensures that attention is paid to minority ethnic, cultural and religious perspectives.This is all the more important where sex equality and minority rights (e.g.freedom of religion) may be on a collision course.If anything, human rights standards concerning equality and non-discrimination are a major site of struggle over the legal acceptability of culturally or religiously inspired practices of sex segregation.

The anti-domination or anti-subordination approach
The anti-subordination or anti-domination approach which we use as an inspiration for this contribution builds on critical legal studies from both sides of the Atlantic.Notwithstanding major differences in the way different authors develop and specify this approach, they show remarkable similarities in their overall analysis, which provides a common ground for addressing issues of equality and non-discrimination. 3Whether they speak in terms of achieving full citizenship for all, of eradicating differences in power, of realising true inclusion or of ending structural disadvantage, to us their analysis boils down to this: though many doubt whether this will ever be accepted in actual legal practice, all seem to agree that the ultimate or true goal of legal guarantees of equality and non-discrimination is the eradication of the subordinate status of historically and structurally disadvantaged groups and ending their domination by the 'upperdogs'.
The focus on ending subordination and domination begs the question how to measure this: which indicators are relevant?Our (limited) literature survey brings out many factors, which we have grouped in three clusters. 4The first cluster of factors focuses on the overall social and economic position of the subordinated group in society, i.e. on the material harm suffered: poverty and a weak economic position, low social status and limited access to education, health care and other facilities and the like are indicators of subordination.The second cluster concerns levels of autonomy and control or power: subordinated groups have little political or economic power to change things; they have less control over their own lives and are more often victims of violence and coercion by others.The third cluster deals with immaterial, ideological factors: subordinated groups suffer from pervasive prejudice, stereotyping and stigmatising that set them aside as less worthy 'others'.
The anti-subordination or anti-domination approach has two important implications for equality and non-discrimination issues, such as segregation.Firstly, it implies that non-discrimination is not to be equated with non-distinguishing, that is with absolutely colour or sex-blind approaches.Even if classifications on the basis of sex or race as such may on the whole be detrimental to women or ethnic minorities, they are not necessarily always unacceptable.If they have a remedial purpose and actually enhance the position of the subordinated group they could be acceptable.Secondly, this approach implies that race or sex-neutral policies, which do not classify explicitly on the basis of race or sex, may still infringe equality or non-discrimination provisions if they negatively affect the position of the underdog group.Again, impact in terms of diminishing or enhancing subordination and dominance is the determining factor.

US case law on race and sex segregation
As far as race and sex segregation are concerned, two landmark cases stand out in the case law of the US Supreme Court.Brown v. Board of Education (1954) on race segregation is no doubt one of the most important decisions in the history of the US and is famous far beyond its borders. 5Less famous, but nevertheless also a landmark case, is United States v. Virginia Military Institute (1996) on sex segregation.

Brown v. Board of Education, 347 US 483 (1954)
In Brown the Supreme Court addressed the claims of black children to be admitted to white, segregated public schools.They invoked the constitutional equality clause to strike down segregation permitted or required under the laws of the states involved.In this historical decision the Supreme Court changed its position as expressed in 1896 in Plessy v. Ferguson that 'separate but equal' facilities for whites and non-whites are constitutionally acceptable and do not infringe the equality clause. 6he Supreme Court considered that even if physical facilities and other 'tangible' factors (such as school buildings, the curriculum or the quality of teachers) may be equal, segregation is not equal as regards all kinds of 'intangible' factors that are difficult to measure.In this respect the Court concluded that segregating black schoolchildren solely on the basis of their race 'generates a feeling of inferiority as to their status in the community that may affect their hearts and minds in a way unlikely ever to be undone (…).We conclude that in the field of education the doctrine of "separate but equal" has no place.Separate educational facilities are inherently unequal'. 7hough the Supreme Court in Brown formally limited this ruling to education, in subsequent cases it became clear that this approach was much more encompassing.Specific rulings referring to Brown struck down segregation regarding public beaches, bathhouses, municipal golf courses, athletics contests, airport restaurants, courtroom seating and municipal auditoriums. 8In the US context, since Brown racial segregation as such is thus regarded as 'inherently unequal'.
The Brown decision focused on the question whether segregation enhances the subordinate position of one group vis-à-vis another group.As James Nickel concludes 'An adequate diagnosis of what was wrong with segregation needs to explain its asymmetrical character: how it deeply wronged African-Americans while hurting whites to a much smaller degree'. 9The crucial argument for striking down the acceptability of 'separate but equal' in Brown regards the question whether the system asserts inferiority.In this respect, the Supreme Court emphasized that 'separate but equal' cannot be considered equal unless both tangible factors such as school buildings or curricula (as was acknowledged before Brown), and intangible factors (such as opportunities and psychological effects) are equal. 10n fact, having become increasingly uncomfortable with the separate but equal doctrine, the Court apparently came to include these 'intangible' factors in its assessment precisely because some states had started to upgrade all 'tangible' facilities for blacks to save their segregation policies. 11At the same time, by taking into account intangible factors such as psychological effects, the Court clearly recognized the relevance of the larger impact of segregation far beyond the specific school context.This means more generally that any assessment of the acceptability of a segregated practice cannot be divorced from the impact and meaning it has in the larger context of society as a whole. 12ited States v. Virginia, 518 US 515 (1996)  Virginia Military Institute (VMI), an all-male public academy founded in 1839, offered training for the 'citizen soldier', that is men prepared for leadership in civilian life and military service.This prestigious institute excluded women.It was financially supported by the State of Virginia.The US sued Virginia for infringing the constitutional equality clause.To justify its male-only admission policy, Virginia argued that VMI could not offer the same sort of education to women without modification, as the male academy was based on an 'adversative method', that is a physically and mentally tough method resembling elite-corps training in the military.VMI considered this unsuitable for women, even if exceptional women could perhaps fulfil the extreme physical and other requirements.To protect VMI's male-only admissions policy, Virginia proposed to develop an equivalent programme for women, the Virginia Women's Institute for Leadership (VWIL), based on a different, cooperative method which would reinforce self-esteem.
The Supreme Court decided against Virginia and agreed with the US that the constitutional requirement of equal protection precluded Virginia from reserving exclusively to men the unique educational opportunity which VMI afforded.The Court considered that the state 'must not rely on overbroad generalizations about the different talents, capacities, or preferences of males and females'. 13It was not disputed that some women, if only a few, would be able and willing to meet the tough requirements for admission.The Court further remarked that 'inherent differences' between men and women should be a cause for celebration, not for denigration or 'for artificial constraints on an individual's opportunity.Sex classifications may be used to compensate women "for particular economic disabilities [they have] suffered" (…) , to "promote equal employment opportunity" (…) , to advance full development of the talent and capacities of our Nation's people.But such classifications may not be used, as they once were, to create or perpetuate the legal, social, and economic inferiority of women'. 14Emphasis added, references to case law omitted.)Turning to the specific situation at the VMI, the Supreme Court emphasized the unique education VMI offered to its male students.In comparison, the proposed VWIL was considered to be a 'pale shadow' of the VMI.Not just because of the envisaged inferior resources, faculty or course offerings available for the VWIL, but especially because of a myriad of intangible factors, such as VMI's 157-year history, its prestige and its influential alumni network.
Notwithstanding this clear conclusion, the Supreme Court explicitly reserved judgment on the question whether 'separate but equal' sex segregation could ever be constitutional. 15This in fact suggests that the Supreme Court allows room for sex segregation if these can be regarded as equal in respect of all relevant tangible and intangible factors. 16

Acceptability of sex segregation: the inferiority test
The parallel between the anti-domination or anti-subordination approach in critical legal studies and the US Supreme Court case law on race and sex segregation is obvious. 17For both, the ultimate test for deciding whether segregation infringes the principle of equality and nondiscrimination would be whether it asserts the inferiority of a group of people or not.A difference, however, lies in the focus of such an enquiry: whereas the anti-subordination approach seems to focus on the macro level of the larger society, the Supreme Court scrutinizes the micro level of the concrete practice itself.Yet, by emphasizing the importance of the impact of all kinds of intangible factors, the Supreme Court links both levels.We would suggest doing so explicitly by combining the insights from both sources.This brings us to formulating the following test to apply to sex-segregation cases.To be acceptable in terms of equality and non-discrimination, any state-endorsed or state-supported policy or practice of sex segregation has to meet a two-pronged test: 1.
Are the segregated facilities or practices genuinely equal as such, that is in terms of all relevant tangible factors?If not, the segregated practice will be unacceptable to start with, though in many cases, of course, this can be rectified by upgrading the inferior separate facilities.If the facilities are equal as far as the tangible factors are concerned, the second test would be: 2.
Do the segregated facilities or practices create or perpetuate the subordinate position of women instead of diminishing their dominance by men?
In this context the three clusters of factors indicating inferiority as identified in anti-domination approaches should be taken into account.The impact of the segregated facilities or practices must be assessed in terms of their wider impact on a) the social and economic position of women, b) their power and autonomy, and c) how women are perceived as the less worthy 'other' (for instance through gender stereotyping and gender role confinement) and whether the hierarchical relationship between men and women is confirmed or challenged.
As non-segregation is the proper default position and all three factors are arguably equally important, it follows that they have to be regarded as cumulative.This means that sex-segregated practices that have a negative impact on women in respect of any of these three factors are not acceptable.It also means that it is up to the proponent of a specific sex-segregated facility, to prove the positive impact of that practice on each of these three factors.
Though space does not allow us to go into this subject here, it should be noted that the focus on the actual (negative) impact of any sex-based distinction in our inferiority test is also a determining factor in equality and non-discrimination review by the Canadian Supreme Court and the South African Constitutional Court.Both courts regard redressing pre-existing disadvantage, stereotyping and inferiority as the quintessence of the constitutional guarantee of equality and non-discrimination. 18n the following section we will explore how the test works out in assessing the acceptability of the three topical issues of sex segregation mentioned in the introduction.All cases involve some kind of state support.Each case study starts with a brief description of the issue, summarizes the main arguments for and against sex segregation we encountered in our research on the topic, applies the inferiority test and formulates some preliminary conclusions.

Introduction
In 1995 fifteen-year-old Nicole Delies, backed up by her club, sued the Royal Netherlands Football Association (KNVB). 19She claimed that she had been treated wrongfully by the KNVB because it did not allow Delies to play with her football team in the national youth competition, just because she was a girl.According to Delies, the KNVB prevented her from developing her footballing skills.Her club claimed that the KNVB prevented them from sending their best players to the competition.Despite the KNVB lawyer's argument that 'football is by nature unsuitable to be played in mixed teams', Delies won her case.The KNVB was ordered to reconsider its policies, which it accordingly did.The case significantly furthered mixed-team matches in Dutch organised football competitions. 20n the Netherlands, as of 2000/2001 mixed teams are allowed in youth football up to the age of 19 at all levels.Up to the age of 12 mixed teams are the rule.From the age of 13, girls may continue to play in a mixed team or join an all-girls team. 21Boys are not allowed to join a girls' team.Adult teams are standard single sex.Professional football remains strictly segregated. 22here is a tremendous amount of (non-legal) literature on gender and sports. 23In an attempt to limit this exploration of the field, the focus will be on amateur football.Thus, issues which are characteristic for professional football are largely left undiscussed.Likewise, issues of representation among referees, coaches and in administrative and managerial functions are omitted.

Arguments for and against (government support for) sex-segregated football
Sex segregation in the male-dominated field of sports, and in particular in sports perceived as 'masculine sports' such as football, is very much a matter of course.McDonagh & Pappano point out that of the more than 500 court cases in the US on sex discrimination in sports, a mere 10% concern practices of sex segregation, not because sex segregation is rare, but because it is so much taken for granted. 24This explains one difference between the case of sports and, in particular, the case of welfare services discussed in this contribution.In sports it is generally not separate facilities that are called for.It is the exclusiveness of male sports that is contested.
The fact that separate play is so common, may also help to explain a second characteristic of the debate on segregated football, namely that opposition to the possibility of separate play as such, at least to giving women a choice between mixed and separate play, are very rare if not absent.The dividing line runs between opponents of mixed play as such and those in favour of free choice (for women).
There seem to be three major arguments underlying the practice of sex-segregation: decency and morals, physical differences and equal rights.

Decency and morals
Although history shows that ideas on what is and what is not appropriate for women (and men) have had and continue to have a very strong influence on women's participation in sport, in the Netherlands such arguments are currently rarely invoked.That is not to say that such arguments, at times linked to religious conviction, do not play a role.It is conceivable that such convictions do influence women's decisions to join a mixed or a single-sex team. 25However, since single-sex teams are still the rule rather than the exception for women over the age of 12, and since the demand for separate facilities only pops up with adolescence, current practice is not problematic for those who prefer segregated sports.Today, similar objections seem to be framed more in terms of physical suitability, an argument that will be discussed hereafter.Therefore, the decency and morals argument will not be further explored here.

Physical differences -real or perceived
Average differences in strength, speed, stamina and so on seem to function as the first and the last word in favour of segregation.For some, it is obvious that women's bodies are just not made for playing football, or at least not for 'real' football.A leading Dutch football journalist wrote in a column on the women's league in 2009: 'There is something strange going on.I suddenly have to like women's football (...).Not even so very long ago, you just went to women's football if you wanted to have a good laugh.(...) It is unnatural to watch a sport you love, being performed by handicapped people.Women's limitations: they can't head [the ball] (...), they are too humble and too sweet.It is like wheelchair basketball.Admirable but it has nothing to do with basketball.' 26 Such opinions on women's football (and sports more generally) are regularly voiced in the media. 27It is not just men saying this.Many women share these feelings and even female top athletes feel 'they don't belong' in the men's competition. 28dolescent girls turn away from sports generally, and mixed teams in particular, apparently not so much because of the machismo and harassment, but because sport no longer appeals to them. 29Girls may also prefer girls-only teams because of real or perceived physical differences, such as a different playing style. 30The assumption that their bodies make women vulnerable, prompts the wish to protect women against injuries, which is another reason to favour separate teams.Male athletes, as well as their supporters, sometimes argue that mixed competition is unfair, because of the chivalrous attitude expected of them.Finally, the idea that women are physically inferior may even be cherished and defended in the face of contradicting facts, by those who fear that women actually beating men in anything but 'female' sports will undermine gender roles. 31owever, despite the apparent self-evidence of physical differences between the sexes, the idea that average physical differences dictate segregation in sports is not uncontested.One line of critique is that the perceived differences in the way men and women move -'throwing like a girl' -are not caused by any physical differences, but are learned.Young argues that women are actively encouraged to move in a 'feminine' way, that is, different from men, with sometimes awkward and ineffective movements as a result. 32Separate play will thus mainly confirm the idea that differences do exist.This effect is enhanced when the rules of the game are specifically adjusted for women's competitions, as is the practice in many different sports.These modified rules, such as shorter matches, lighter balls, smaller fields and the like, do not seem to serve any particular purpose and do not appear proportionate to any real or perceived physical differences. 33nother familiar argument is that far greater differences exist between individual women and between individual men, than between women and men as groups.This is illustrated by female top athletes who are able to play 'with the boys' and maybe even win, but are still excluded because they are women. 34lso, it has been pointed out that in some team sports, such as American football, athletes perform different functions: some players must be bulky blockers, whereas others need to be sprinters.This latter argument, of course, is not so much an argument against segregation as it is a rebuttal of the inevitability of single-sex teams.

Equal rights arguments
Discussions on sports and equal rights focus nearly as much on physical differences as do the discussions presented in the previous section.The arguments discussed in this section differ from those discussed above, in that they start from the presumption that somehow, equality should be achieved.The discussion revolves around the question of what, exactly, does equality in sport look like, and whether it is best achieved by separate or mixed play.
Equality-based arguments supporting segregated play mostly stress the importance of (the possibility to choose for) separate facilities for women, not men, and only for those sports where significant and relevant average physical differences exist.Such an equality approach has been codified, for example, in the US, in Title IX of the Education Amendments of 1972, albeit in a symmetrical way.
Demands for women-only teams based on sex-equality arguments may regard the opportunity to practice a particular sport as such, 35 the possibility to benefit from government funding in support of sports, 36 the possibility to develop one's skills without harassment by jolly male peers et cetera.Two arguments seem to go to the heart of the matter: the first is that many girls and women for whatever reason simply do not want to play with men, and they will drop out when they are coerced into mixed teams. 37The second argument is that if women are forced to play with men, they will always remain second rate, because of superior male performance. 38he 'Title IX approach', however, has also raised significant criticism. 39This critique is partly due to ways of implementation that do not serve the cause of equality at all, such as considering field hockey to be a suitable feminine equivalent to male baseball or football, forcing girls into girls' things and boys into boys' things, thus enhancing instead of diminishing equality.
McDonagh & Pappano argue that starting from the idea of physical difference is not innocent.In the context of sports a difference is translated automatically into women's inferiority, or worse, into women's natural inferiority. 40The implication of this argument is that segregated play will automatically strengthen differences and is thus detrimental to women.Nevertheless, McDonagh & Pappano do not advocate the abolition of the possibility of women-only teams; they 'merely' wish to end male exclusiveness in sport.
Another equality-based argument against segregated play is that there seems to be no good reason to grant women their own league: if the performance of female athletes is inferior to that of male athletes, that is, if they are unable to really compete with men, how could equal treatment, in the sense of spending equal amounts of money, et cetera, ever be justified in the name of equality? 41Elling & Knoppers have countered this argument by pointing out that there are other factors that influence people's interest in sports.It is not just the fastest and the strongest athletes that can count on public and media attention.In the Dutch primary (football) league, for example, most clubs are unable to really compete with the top clubs.Nevertheless, all matches are followed with eager interest.Aspects such as being familiar with the athletes, close competition, and geographical links (local, national) are equally responsible for arousing interest. 42 very different kind of argument that is nevertheless equality-based is that women want to have it both ways: this boils down to the idea that women have to choose: either they compete with men (and lose, so, apparently, is the assumption) or they stick to their own separate teams and competition.This kind of argument does not do well in view of the historical subordination of women inferiority test.

Equal or unequal facilities
Research as well as practice show that facilities for women's teams are not of the same standard as those for male teams.They have to practice at odd hours, on poorly graded fields, and get the least qualified or able referees and coaches. 43Even where there is a legal requirement to spend (relatively speaking) as much money on women as on men, as is the case in the US, in practice still huge differences exist. 44oreover, differences in the numbers of participating women and men, as well as differences in the kinds of sports attracting women and men, cause problems.For instance, the fact that relatively few women play football, implies that team-mates will vary more in age, that they will have to travel farther to meet their adversaries, and that there is a chance that their league has fewer matches than the men's league.This kind of problem is not easily solved.
However, at least in cases where men and women belong to the same club, it should be possible to measure and guarantee equality in possibilities and facilities, in terms of the fields teams may use, the designation of referees and coaches and so on.Thus, although generally, separate women's teams are far less well off than men's teams, even within the same club, arguably that does not necessarily have to be so.'Given the proper opportunities and exposure, women have been extremely successful in sports like basketball and soccer.' 45 That means that if sex-segregated play would be considered beneficial for women, for instance because of its empowering potential (see below), governments should make sure that facilities for women are (at least) as good as they are for men.

Perpetuating or diminishing subordination Social and economic position
There has been a rapid increase in women's participation in sports, even in typically 'male' sports such as football.This is generally regarded as a welcome trend.Sport is considered to be healthy, good for one's character and body and, moreover, sport is a major field of social life, where enormous sums of money circulate.The importance of sport, and in particular football, for social life, is difficult to underestimate.In the Netherlands, for example, the Monday newspaper always consists of two parts: one half contains general news, the other is dedicated to sport, and half of that to football.Thus, arguably, participation in sport will generally have some influence on people's social and economic position, although effects will differ, depending, for instance, on the level at which sport is practised. 46However, it is less clear whether and how segregation influences women's social and economic position.In some sports, such as tennis, the women's circuit has attained popularity levels comparable with the men's tournaments.Hardly any of these female athletes could, in all probability, have reached the same level of popularity, fame and media exposure if there had not been a separate circuit.Also the fact that many girls seem to drop out if they cannot enter a single-sex team once they reach adolescence seems to be an indication that separate facilities for women are the better road to enhance women's social and economic status in sport.At the same time, the higher levels of all kinds of sports remain hermetically sealed for (capable) women, thus excluding them from the top salaries and other profits that are still mainly the privilege of male athletes.It is clear that women's competitions are generally far less profitable than men's.One only needs to look at the incredible sums of money earned by the world's (male) top footballers.
Generally, participation in amateur sports will not raise the athlete's social or economic status that much, although participation at the higher levels will doubtless generate certain benefits.Research indicates a link between the cost of a ticket -which is generally significantly higher for men's matches -and the evaluation of women's and men's teams. 47However, whether this should lead to the conclusion that mixed play is better for women's social and economic position, or that the price of tickets for women's games must be increased (which entails the risk that the public decides not to show up at all), remains subject to debate.
Arguably, separate play potentially benefits many women, whereas access to the male (top) leagues will be profitable for a few women only.This arguments offers support for the idea that women should have a choice.At the same time, and in light of the mixed blessings of equaltreatment legislation in sport such as the American Title IX approach, a close watch should be kept on all valuation of men's and women's matches et cetera in terms of money: somehow, a balance must be struck. 48

Power and autonomy
As in the other two cases discussed in this paper, participation as such is considered to be important. 49Whether one derives satisfaction and self-esteem by participating in sport may, however, very well be dependent on one's skills.Good athletes may feel empowered by the esteem they gain with their performance.Less skilled athletes may experience this differently, in particular if their participation is not (entirely) voluntary, such as in school or because of overbearing parents.Arguably, girls and women who do not possess notable athletic skills, but still enjoy the game, may be better off in a single-sex team, because they feel more confident.
Again, the male exclusiveness of the higher sports levels deprives the better qualified female athletes of the possibility to develop their skills to the best of their ability.Just like Delies, they may demand access to male teams because they want to play at the top or they want to compete with equally skilled athletes.

'Othering'
It has regularly been argued that sport is 'a site where ideological beliefs about gender are deeply entrenched'. 50At the same time, sport may also contest the construction of masculinities. 5127 History shows that women's participation in sport is bound to influence dominant ideas on what is and what is not appropriate behaviour for women.On the other hand, as has been pointed out by quite a number of authors, even many of the most successful female athletes participating in 'male' sports such as football, feel compelled to stress their femininity, thus stressing their difference from 'the other', that is men.Outstanding female athletes run the risk of being accused of being male (or at least: not female).Recently, South African runner Caster Semenya was asked by a journalist, after an impressive accomplishment, whether she was a man.The International Association of Athletics Federations wondered as well, and so her glorious victory ended in a prolonged and humiliating investigation into her 'true sex'. 52In order to avoid these kinds of problems, female athletes grow their hair long, wear jewelry, regularly show up in evening dress, use make up and nail varnish.Although difference in itself is not 'bad' per se, it has been pointed out that the sexualisation of women athletes trivializes their performances, thus conveying a message of women's inferiority. 53McDonagh & Pappano conclude that the essential point of this stress on femininity is that these athletes can thus still be seen as 'girls': '[w]omen can play sports as long as their participation reinforces, rather than challenges, the view of women as heterosexually attractive and lesser athletes than men.They may be good or talented, but such recognition must exist within the scope and limits of female play.' 54omen playing in their own leagues seem inclined to stress their femininity more than the few women who have made it so far to a male (top) team.Explanations that come up are that a woman playing in a male team benefits more from being tough than from being elegant.Also, the few women making it to the top might care less about being mistaken for a man than women playing in their own league.A male-looking woman participating in a competition where athletes are solely measured according to performance does not need to be sex-tested, because her sex is irrelevant for the valuation of her performance.The same woman competing in the women's league, however, may be suspected of really being a man, and thus of foul play, justifying sex tests, as in the case of Semenya. 55lso, it sometimes seems as if women play a different game than men, even when they are practising the same sport.This, of course, links up to the 'feminine way' of moving, as discussed earlier.Joseph Blatter, the Chairman of the highest authority in football, FIFA, is reported to have said: 'Female players are pretty, if you excuse me for saying so, and they already have some different rules to men -such as playing with a lighter ball.That decision was taken to create a more female aesthetic, so why not do it in fashion?' 56 If this is what segregated play comes down to, 'othering' and gender role confinement will indeed be very much enhanced.
Finally, keeping (top) male competitions closed to women arguably makes for the worst kind of 'othering': after all, the main reason for barring women seems to be that women are just not good enough.So, if female athletes, even the best among them, are forced to compete among themselves, that quite automatically sends the message identified by McDonagh & Pappano that women are 'naturally inferior'. 578 Cf.Elling 1999, supra note 20, p. 27, who remarks that women's football needs more support and appreciation from their own ranks, good international performance, and lots of media coverage, to improve its image.

Preliminary conclusions
Inequalities in the facilities offered are widespread and many.However, such inequalities cannot be a sufficient reason to deny women the opportunity to play in single-sex teams, if one accepts the widely supported view that women should be able to choose for themselves whether to join a mixed team or play in a single-sex team, at least in the present context where sports are so thoroughly gender-biased in favour of men, and in light of women's historical subordination.Rather, governments should make sure that their money is not largely spent on men, to the detriment of women's sports facilities. 58ith regard to the effects of sex segregation in sports in terms of social and economic position, power and autonomy, as well as 'othering', the test provides good arguments to support the conclusion that, at least for now, sports, and football in particular, is such a masculine environment that forcing women to play in mixed teams is not a good idea.Most women will just drop out and miss out on everything that may be gained through participating.
However, to reduce the detrimental effects of othering, in particular gender role confinement and messages of inferiority, as well as to give women the best opportunity to develop their skills, male exclusiveness at the higher levels has to stop.Also, the different rules for many women's leagues, such as playing with a lighter ball, shorter matches, and so on, have to be dispensed with.They do not seem to serve any particular purpose, whereas their effect in terms of emphasizing the inferior skills of women -women being weaker, having less staying poweris considerable.
The inferiority test thus leads to the conclusion that governments should both support a free choice for women, and refuse to support exclusive men's teams.In accepting such segregation, they actively support sex discrimination.Capable women must be allowed to join, and should not be barred because they happen to be female.

Introduction
In European countries single-sex primary and secondary education have been abolished for several decades or have decreased considerably. 59In the Netherlands, for example, there are no longer any single-sex schools since the disappearance of special secondary schools for girls (the MMS) and the Catholic boys' and girls' schools in 1968.In other -formerly predominantly Catholic -European countries similar changes took place more recently. 60In Anglo-American countries like the UK and the US single-sex (boarding and day) primary and secondary schools are still quite common, although they have notably diminished during the last few decades and are mostly of a private or independent character. 61owever, sex-segregated primary and secondary education is occupying centre stage once again, in some instances even supported by the state.The numbers are small, but two issues are telling.Firstly, in 2001 the first all-girls Muslim secondary school became a state school in the UK. 62There are also Muslim boys' schools in the UK, but these are private, not state-funded.Most Muslim schools in the UK are mixed. 63Sometimes mixed Muslim schools also employ practices of sex segregation.In the Netherlands there are reports of separate entrances, male and female areas within the same building, (sometimes) single-sex classes or -at a more modest level -girls' and boys' rows in mixed classes. 64econdly, in 2002 the No Child Left Behind Act and a 2006 Bush Administration regulation supported a 'resegregation' of public schools in the US, facilitating pedagogical experiments including single-gender classrooms and supporting voluntary sex segregation financially as long as 'substantially equal' opportunities are also provided for the excluded sex and when it meets an 'important governmental objective'. 65This objective includes the improvement of educational achievement, through the provision of diverse educational opportunities and the meeting of particular educational needs of students. 6667 The two issues mentioned above give rise to two major arguments in favour of governmentsupported sex-segregated education.These arguments focus on the effectiveness of education and on respect for religious practices.It will become clear that behind these arguments lie concerns about equality: are differences highlighted to affirm or to overcome dominant and subordinate gender roles?These arguments are thus directly related to the inferiority test which will be applied in Section 4.3.

Effectiveness of education
The effectiveness of education is generally defined in terms of the improvement of performance and/or the strengthening of self-esteem of students.We will start with two arguments about effectiveness in favour of single-sex schools, that do not refer specifically to either boys' or girls' schools.

Arguments for
The first argument is that single-sex schools or classes positively influence the concentration of pupils, compared to the hormonal distraction of pupils sitting together in mixed classes. 68he second argument for single-sex schools or classrooms, especially put forward in the context of the 'resegregation' in the US, says that sex segregation, in addition to co-education, provides 'diversity' in educational options and increases the choice of parents to find the school that fits their children best. 69This seems to be especially relevant for low-income minority groups, which cannot afford single-sex schools that are mostly of a private nature. 70n contrast to these arguments for single-sex schools, irrespective of which sex it concerns, most arguments focus specifically on either boys' or girls' schools.

Boys' schools
The primary argument for the renewed introduction of boys' schools is to solve the 'boy crisis'.This refers to the development that boys are falling behind girls in their educational performances. 71ne strand of this argument considers boys to be victims of the feminization of education.This feminization has disqualified their masculinity, and the focus on the advancement of girls has come at the expense of boys' self-esteem and success.Another strand refers to boys as victims of an education system that does not take sufficiently into account the sequential developmental differences of brains and sexuality.Put differently and more generally: boys and girls are biologically different, so it is only 'natural' to take this into account through single-sex schools. 72oth strands in favour of boys' schools seem to stem from a conservative motive that stresses 'natural' differences with the aim to restore the conditions for the privileged position of men in society.

Girls' schools
One set of relatively 'new' arguments for girls' schools or classes focuses firstly on protection against gender discrimination and secondly on empowerment.Protection against gender discrimination entails protection against sex stereotyping, and fewer learning and interaction opportunities for girls in mixed classes. 73Empowerment is the goal of girls-only education in traditionally male topics, like science and maths.The aim is to overcome prejudices and gender stereotypical behaviour in order to reach higher self-esteem, better results and increased study choices among girls with regard to these 'male' topics, which could also enhance girls' job opportunities.
The motive seems to stem from a progressive view: separate education for girls is advocated to compensate for unequal opportunities in education and to enhance equality.

131
Not clearly progressive and mostly put forward as a justification for girls' classes or schools is the argument -already playing a role in the case US v. Virginia (Section 2) -that boys and girls have different learning styles. 75guments against the effectiveness of single-sex schooling Against the diversity and choice argument is said that these values are not inherently good.Diversity could lead to separation and encourage social fragmentation, and choice could occur at the expense of equality. 76he hormonal distraction argument is not seriously countered, probably because this is not considered to be a problem.If there is a problem of concentration, this is mainly explained by boys' disruptive behaviour in general, which will not be solved by single-sex schools. 77ther arguments against boys' and/or girls' schools or classes have the following two strands.Firstly, there is no clear empirical evidence of problems of mixed schooling and successes of single-sex education.Secondly, sex segregation is no solution for the problems in mixed schools or classes.

No empirical evidence
The first strand in the opposition against single-sex schools is that scientific research has not revealed a 'boy crisis'.Some contest that there is a 'crisis' or downplay the perceived 'gender gap', often pointing to other variables than sex.School performance, so it is said, is mainly determined by the ability and social background of pupils, and after that by school factors, such as leadership and teacher expertise. 78Socialization is mentioned, that stimulates boys to be physically dynamic and playful and girls to be quiet and serious and, as a consequence, benefits the educational successes of girls. 79Boys and girls from similar backgrounds generally have similar success, although gender differences are especially likely to be found among economically disadvantaged children.So, biological differences are there, but they are overestimated.Especially socio-economic background is mentioned as a variable of more influence than biology: particularly low income minority boys perform worse.They obtain lower grades, have behavioural problems and high drop-out rates.Instead of 'failed boys' or a 'boy crisis', there seems to be a 'minority boy crisis', which in the US concerns primarily African American boys. 80n general there is also no clear evidence that single-sex schools perform better than coeducational schools.Boys' and girls' schools vary too much and are too few to make tenable comparisons and firm conclusions.At best there are some modest findings, but these are far from the often claimed success stories in the popular media. 81urthermore, there is no clear evidence that girls in all-girls' schools or classes have higher self-esteem, make less traditional subject choices or do better in mathematics.

132
that girls in single-sex schools have less stereotypical views of maths.Interestingly, this goes together with 'magnified stereotypes of the subject' among boys in single-sex schools. 82However, this last result is contested by others, who maintain that particularly in mixed school settings subjects like physics and drama become more easily gendered, suggesting that single-sex schools offer less gendered opportunities to subject choices. 83x-segregated education is no solution for problems of co-education The second line of thought is that single-sex education is not a good solution for the problems associated with mixed schooling.Boys most likely do not flourish in single-sex environments, because the distraction of the all-male peer group could have negative effects on learning.However, some research seems to indicate that boys' schools could be beneficial for disadvantaged minority boys. 84robably girls could gain from single-sex education, but maybe also the opposite of the empowerment of girls is the result, as single-sex education tends to strengthen attitudes that favour traditional, even stereotypical, views of gender roles.In this way, sex segregation in education could have the aim of enhancing equality, but turn out to advance inequality.Furthermore, sex-segregated education offers no real protection against sexual harassment.Instead of trying to combat attitudes that could lead to sexual harassment, boys and girls are simply put apart.As far as problems occur in mixed education, one should work on solutions in that context, identifying the key components of a good educational environment, especially in inclusive, varied and gender-sensitive teaching methodology. 85

Freedom of religion
The second major argument for single-sex schools stems from religious considerations.These are not really new: Catholics have had a long tradition of single-sex schools.More orthodox Jews in Western Europe upheld sex-segregated practices in education at a time when demands for gender equality and secularization led to the rapid decline of single-sex schools.However, more recently, religious arguments have been re-imported by Muslim immigrants.In the following the focus will be on Muslim single-sex schooling as a topical multicultural issue in Western Europe.

Arguments for
The norm that the sexes have to remain separate (outside the context of the extended family) from the age of puberty is the main argument for single-sex schooling among Muslim parents.Closely related to this religious argument are concerns about the sexual (and racial) harassment of Muslim girls, concerns about 'apparently lax sexual mores in society generally, and in secondary schools in particular', clothing issues related to modesty (especially for physical education and swimming lessons) and sex education.Apparently these arguments are especially voiced in favour of Muslim girls' schools. 86he difference between Islamic arguments and secular arguments for single-sex schools is that the former stresses sexual mores (of girls) and the latter's educational performance.87 K. Haw, 'Muslim Girls' schools -A conflict of interests?',1994 Gender & Education, no. 1, pp. 70-74.88 English, supra note 66, p. 27 (referring to the Department of Education) and Levit, supra note 68, pp. 7, 12 (referring to the anti-empiricism of American courts).89 One research carried out in Canada suggests that girls in all-girls private schools would benefit, but not girls in a public same-sex school.

Arguments against
Interestingly, most arguments against Muslim girls' schools are directed primarily at religious schools in general or Muslim schools as such.Muslim girls' schools seem to be perceived as an extreme type of religious or Muslim schools.In those arguments two concerns can be distinguished.
The first concern is about equality between men and women.Opponents of religious or Muslim schools are afraid that girls will be taught to take on a subordinate gender role and will be prepared for marriage and the household.The second argument is about separatism.While the ethnic, cultural and economic divide between Muslims and non-Muslims is already alarmingly deep, special education for Muslims, and for Muslim girls in particular, could broaden the gap with the dominant society.
Some arguments against state-funded Muslim (girls') schools try to find the solution in a greater attentiveness for the needs of Muslim boys and girls in general state-funded schools, in order to discourage the establishment of separate schools for Muslims.This could include antiracist and anti-sexist policies, appointing more teachers with the same ethnic, cultural or religious background, girls-only classes in some subject areas or even the retention of non-Muslim singlesex schooling. 87n the next section we will see what these arguments mean for the application of the inferiority test and specifically whether the inferiority test is able to deal with ethnic, cultural and religious aspects of single-sex education.

Equal or unequal facilities
Equality of tangible factors does not receive any attention in the literature on Muslim girls' schools in the British context.This lack of attention could be explained most likely by the strict requirements (as to curriculum, housing facilities etc.) to enter the state-school system in Great Britain, as would be the case in the Netherlands.
In the US equality of tangible factors of sex-segregated education is of greater concern, due to the absence of strict supervision on equal provisions in sex-segregated programmes. 88

Social and economic position
If girls-only education improves school achievements and strengthens the self-esteem of girls, also in traditionally male subjects, it will most likely have a positive impact on their social and economic position.Although there are some indications along this line, empirical evidence does not unambiguously reveal that single-sex education has this effectiveness impact on girls in general.Maybe (some) girls could benefit from girls' classes in traditional male subjects like maths, science and ICT. 89Until now there is no clear evidence that this results in choices for vocational education in these areas and subsequent jobs with social and economic standing.
Girls-only as well as all-boys schooling, whether religious or non-religious, could even be detrimental to the social and economic position of girls, as far as these confirm traditional gender roles.This could perpetuate the subordinate position of women and the dominant position of men, considering the lower status of 'women's' jobs and financial dependence in case of motherhood et cetera.

Power and autonomy
Sex-segregated education was introduced in the US as an additional option for parents to meet the educational needs of their children.The problem, however, is that sex-segregated education may not always extend options; it could also lead to a reduction in co-educational opportunities.
Religious single-sex schools could be seen as an option for parents to ensure that their child is educated in conformity with their religious beliefs.As far as a girl has a say in the choice between a mixed or single-sex religious (or non-religious) school, this could be seen as strengthening her autonomy.
However, single-sex education is not clearly beneficial for the school performances of girls generally.There is also the tendency of single-sex education, as compared with co-educational settings, to confirm traditional gender roles.This tendency could change into a conviction that traditional gender roles are good in conservative environments, like in traditional cultural or patriarchal religious single-sex schools.So it is also quite likely that Muslim girls' schools reinforce traditional roles.
Maybe one could contend that parents consciously choose for such an education, but it is more difficult to maintain this for a minor girl.
An empowering effect could occur in all-girls schools or classes (in traditional male topics), because of the absence of male dominance and sexual harassment and when they are used to combat existing sex discrimination and to overcome sex stereotypes.The question, however, is whether this empowering effect will last in mixed-sex contexts.

'Othering'
Sex-segregated education singles out gender as a relevant variable for all pupils.That is why sex-segregated schooling is at first sight suspect: it could very well be based on overbroad generalizations and stereotypes of the abilities, characters and prospects of girls and boys.Thus boys' schools that affirm their 'masculine' abilities and girls' schools that aim to protect girls' modesty and enhance 'feminine' capacities seem to entail powerful role confinement.Such gender stereotypes and role confinement imply almost always the perpetuation of the hierarchy between men and women, rendering girls as less worthy others.Muslim girls' schools seem to be scrutinized most strictly, although one may wonder whether this is out of concern for girls or based on (intersectional) gender, racial and or religious stereotyping, thereby affirming the 'othering' of girls, Muslims or Muslim girls.

Preliminary conclusions
What is the outcome of the inferiority test?Does it give guidance to assess the acceptability of state-funded single-sex education?
The first part of the inferiority test could be helpful to assess whether the provision of girlsonly schooling is equal compared to boys' education and co-education.The outcome depends on the facts.
The second part of the inferiority test gives a varied outcome on single-sex schools.Firstly, the outcome on the social and economic position of women as well as their power and autonomy is not really clear, because of the lack of solid empirical evidence of the impact of sex-segregated education on these issues, although there are indications that girls will benefit more from singlesex schools than boys.Secondly, all-girls classes in traditionally 'male' topics that are meant to overcome gender discrimination will pass the inferiority test most easily, despite the weak empirical evidence of the positive effect on the social and economic position of women as well as their power and autonomy.Thirdly, single-sex education that confirms traditional gender roles will not pass the inferiority test.Traditional gender roles will perpetuate the subordinate position of women on all the three elements of the second part of the test: their social and economic position, their power and autonomy and 'othering'.The inferiority test focuses on the equal, subordinate or dominant relations between men and women, which is not surprising as an assessment instrument for sex-segregated practices.However: girls' and boys' schools are sometimes advocated with reference to a subordinate minority ethnic, religious and class position.Is the inferiority test (sufficiently) sensitive to these positions?
As applied above, the inferiority test seems to be indifferent to the Islamic (or other religious) denomination and the ethnic character of girls' schools.The test is about gender relations, with special attention for the effects of single-sex schooling on the vulnerable position of women, irrespective of whether sex segregation is inspired by religion, culture or another factor.Yet, the test shows that religion could play an autonomy-strengthening role (if girls are free to choose for a religious single-sex school) as well as an autonomy-weakening role (as far as religion reinforces traditional gender roles).Minority ethnic, cultural and religious positions can also figure in 'othering'.
Moreover, the inferiority test could pay attention to the class position of boys and girls, while the test includes the element of social and economic position and there is also evidence that social background is a variable that determines school performances.In this context, it is interesting to consider lower-class minority boys.It is hoped that these boys will benefit socioeconomically from all-boys' schools, but as this is sought to be realized through single-sex schools, it seems to confirm the 'othering' and gender domination of boys.After all, the question is why and how 'black' boys' schools are more effective in empowering lower-class and minority boys than co-educated 'black' schools.

Introduction
In 2009 several controversial cases involving sex-segregated welfare services gave rise to considerable debates in the Dutch media.A first case concerns separate civic integration courses (inburgeringscursussen) for men and women belonging to non-Western minorities, organised and/or financially supported by local government.A second issue concerned the city of Utrecht, which attracted a lot of media attention by subsidizing separate information offices for men and women. 90An added bone of contention was its location in a mosque.The offices provided information on issues such as social welfare, education, work, sports and all kinds of municipal Parliamentary Proceedings 136 services.Both segregation practices inspired parliamentary questions being raised and motions passed to put an end to them. 91Due to this pressure they have been or will be discontinued. 92

Arguments for and against (government support for) sex-segregated civic integration courses and other welfare services
The arguments for and against the (local) government supporting this kind of segregated services seem to revolve around three major issues: the accessibility and effectiveness of the services involved, the equality and emancipation of women, and respect for certain religious values and practices. 93

Accessibility and effectiveness
One of the major arguments for supporting segregated services regards accessibility: they are believed to enhance access to outreach facilities for groups which are difficult to get in touch with, such as some groups of minority women.As the responsible alderman for the Utrecht policy argued in the discussion on segregated integration courses: 'We prefer to teach them separately over not teaching them at all'. 94This is all the more important where some women may not be allowed by their husbands to make use of the facilities unless in a segregated setting.
Opponents argue it is questionable whether many women refrain from using the facilities for this reason.Additionally, they fear that the availability of segregated facilities may indeed have the undesirable effect of increasing the number of women choosing segregated facilities (who would otherwise adjust to having mixed-sex facilities).As regards the specific issue of civic integration courses they also point to different solutions to the problem of making it difficult to reach minority women, for instance by facilitating pre-integration projects, providing for childcare, financial incentives or providing integration in the workplace, at school, et cetera.
A second major argument supporting segregated facilities focuses on the better results they are presumed to entail.The targeted women may be more at ease and feel freer to ask questions and share their problems in an all-women environment.In the company of men, women will often remain silent.Regarding the separate civic integration courses it is emphasized that having segregated classes does not mean having a different programme.The programme is the same, but it will be easier to discuss values of sex equality in segregated classes and this will enhance their emancipating effect.As such, segregated facilities may simply be much more effective.Opponents will argue that no evidence exists which really proves such effects and, again, even if so, that segregation need not be the only solution available.Besides, especially if women tend to be intimidated by the presence of men it is all the more important to teach them (and the men) to counter this mechanism.

Sex equality and the emancipation of women
In the public discussions sex equality and the emancipation of women seem to figure as the single most important objection to segregated welfare facilities.Many people perceive sex segregation as inherently 'wrong' and in violation of the principle of sex equality, which is emphasized as fundamental to Dutch society.Sex segregation is easily equated with the subordination of women.It should only be allowed in exceptional and completely acceptable or uncontested situations, such as separate bathrooms or safe houses for battered women.As far as integration courses are concerned this argument is given special pertinence.These courses are meant to familiarize immigrant groups with Dutch values and practices.Whereas sex equality is a fundamental value, segregated classes give the opposite signal.
Related concerns argue that sex segregation reinforces gender role confinement and/or gender stereotypes.Others would say that sex segregation is not a gender-neutral practice of separating both sexes on a basis of equality, but in fact means separating women.(As far as civic integration courses are concerned it is women who are presumed not to be reached by nonsegregated classes; apparently men are presumed not to be discouraged; this is born out by the absence of male-only classes).Separate classes or facilities more generally may thus point to women as being the problem.Still others point to the danger of a sliding slope.Where will it end?Sex-segregated ticket offices for public transport?Separate seats for men and women in public buses, trains and waiting rooms?A last concern we would like to mention here regards the question whether providing sex-segregated facilities will lead to sex-based selection for jobs as a consequence.
The above arguments are contested by those who point to the free choice of the women (and men) concerned: segregated facilities are voluntary, not compulsory.As such, providing both segregated and mixed facilities extends the choices available which should be applauded.Alternatively, providing for segregated facilities can be perceived as a form of preferential treatment for women.
More generally, some people have raised the question of why should we object to practices that were quite common in the Netherlands only several decades ago.All kinds of sex segregation existed until well into the 1970s, yet emancipation came about 'by itself'.Besides, when the second feminist wave was in full swing in the seventies and eighties, exclusively female groups played an important role in consciousness raising and the emancipation of Dutch women.So why worry?Why should such an emancipation process not work for ethnic minority women as well?Others object to this optimistic scenario and point to the specific context of current requests for separate facilities by ethnic minorities as quite different from sex-segregated practices prevalent in Dutch society 50 years ago: the latter segregation was part of a society which shared, at the bottom, more or less the same values, world-view and language.The former, however, is not embedded in society at large (if so, no integration courses would be necessary).

Freedom of religion
The major reason for offering segregated courses stems from the religious and/or cultural objections of some minority groups to mixed-sex facilities.The willingness of certain municipalities to accommodate this enhances respect for religious freedom and is in line with the approach of the Dutch Equal Treatment Commission.In several cases the Commission has considered sexsegregated facilities acceptable and, in some situations, even legally required: a failure to 95 See oordeel 2004-95 and oordeel 2005-86, see <www.cgb.nl>.The judgments of the Commission are not legally binding.Regretfully, they are not available in English.In both cases the issue was whether an organization (in the first case a school, in the second an organization of social counsellors) violates non-discrimination provisions by its policy of refusing to speak to women wearing a face-covering veil.(The argument being good communication.)The Equal Treatment Commission considers the policy to constitute indirect discrimination on the basis of religion as this argument does not meet the required test of an 'objective justification'.Though the Commission acknowledges that the aim of the policy is legitimate, the means do not meet the test of subsidiarity.In both cases the organization could allocate the veil-wearing woman to a female employee, with whom she does not feel obliged to wear the veil.96 See supra, note 92.138 accommodate religiously inspired practices of sex segregation may amount to indirect discrimination on grounds of religion under the Dutch Equal Treatment Act. 95hough freedom of religion is undisputedly a fundamental right; in the media debates the freedom of religion is not put forward as a major argument to support sex-segregated facilities.Surprisingly, neither does the Minister of Integration refer to this fundamental right in his letters to Parliament or the municipalities on this issue. 96

Equal or unequal facilities
As nothing in the debates points in another direction, we will presume that the sex-segregated facilities as such are equal in all relevant tangible aspects.So we will proceed to the second part of the inferiority test.

Perpetuating or diminishing subordination
Social and economic position On the whole it seems evident that making use of the kinds of welfare facilities at issue here will improve the social and economic opportunities of the women involved.Being better informed about all kinds of things will improve their ability to participate in society.As such, any barriers to the accessibility of those facilities should be removed.This is all the more important as most members of the groups involved are particularly vulnerable socially and economically.Thus, if it is established that otherwise (some) women stay away, this first element of the inferiority test would not be a bar to but even in clear support of segregated facilities.
This would be even more so if it were established that in fact women benefit more from segregated than mixed-sex facilities.If they feel more at ease in asking questions without the prying eyes of men, talk more freely about their problems and discuss these with fellow women, it may well be that the facilities provided are much more fruitful.Though quite plausible, whether this is indeed the case is contested.In this respect, social science expertise may hopefully help us out.

Power and autonomy
The question whether segregated facilities will empower women and strengthen autonomy or reduce it seems to lead to a more ambivalent answer.
On the one hand, generally speaking, improving social and economic opportunities will be empowering and beneficial to autonomy.On the other hand, the specific effects may well depend on the actual motives for participation in segregated facilities.Different situations suggest themselves.
If making use of segregated facilities results from the free choice of the woman involved, there does not seem to be any problem in terms of empowerment and autonomy.We should presume that the woman herself knows best what suits her.If it is for religious reasons we may have to accept that the woman involved subscribes to orthodox views on the proper relationship between men and women.Should the freedom of religion not be respected whatever its content unless it infringes the rights and freedoms of others?Besides, the woman concerned may well have other reasons for preferring female-only classes, such as feeling more at ease, simply enjoying the company of women over men in this setting or having her own assessment of what is most emancipating and empowering for herself and/or fellow minority women.
If, however, making use of segregated facilities is enforced by the woman's spouse and/or family/community this is indeed problematic: in that situation providing segregated facilities means giving state backing to the pressure put on women.If the woman involved would not be allowed to make use of the welfare facilities at all unless segregated, the overall empowering effect will probably still be positive.But what if women are involved who prefer making use of mixed-sex facilities and would be allowed to do so unless segregated facilities are available?Such women will lose in terms of empowerment and autonomy by being forced to use segregated facilities.Of course, this raises the crucial question of why and how many women will opt for segregated facilities of their own free will, how many would be forced to do so and how many would be kept away from any facilities unless these are segregated.So far, no clear numbers based on thorough research are available.

'Othering'
This element is probably the most difficult to assess.It is a strongly 'intangible' factor.The following questions would seem relevant.
Firstly, is sex segregation to be regarded as 'wrong' in itself because it is to be equated with subordinating women?Many commentators aver this in different degrees, including the Dutch Minister of Integration.Others would say the women involved can decide that for themselves.Although this may well be the case, would it be determining?What about the external perceptions involved?
Secondly, is gender role confinement enhanced?We would say 'yes'.Regardless of whether at the end of the day separate facilities may be more effective and beneficial for women, at the same time they may very well enhance images of the propriety of separate roles and spheres of living for women and men in society.Images that may send out the message that there is nothing wrong with looking with suspicion at the intermingling of the sexes outside the confinement of the home.This is even more so where sex segregation is expressly motivated by religious views on the proper roles of men and women.Given the historic role of this type of thinking to justify the subjection of women throughout the ages, we should be very wary about its effects in modern times as well.
Thirdly, is sex stereotyping enhanced?Again it seems that this could very well be the case.To start with, sex segregation does not seem to be a gender-neutral practice as it seems that it is women who need to be set apart.Besides, women attending segregated courses seem to be perceived as helpless victims of male oppression, as women without real agency.Not many commentators seem to acknowledge that for at least some women segregated facilities could be a truly free choice.
Yet more important than sex stereotyping, it seems that religious and/or racial stereotyping is at issue: a preference for segregated facilities is perceived by the larger public as a typical manifestation of reprehensible norms and values within Islam and/or non-Western minorities as such.Minority women are perceived as victims of their culture and religion and/or their oppres-sive men.Dutch/Western culture is contrasted with this as a progressive, equality-based culture which abhors any inequalities between the sexes.

Preliminary conclusions
Sex-segregated welfare facilities of the kind discussed above will not pass all the elements of the inferiority test.Especially the elements of 'power and autonomy' and 'othering' raise serious doubts about segregation in this context being acceptable.Yet, in many instances we seem to have to proceed on the basis of the presumed effects of having either segregated or mixed-sex facilities.In this respect it is clear we are in need of much more empirical data to reach better informed conclusions.
As long as more and better information is lacking, the application of the inferiority test would seem to suggest that proceeding with sex-segregated welfare facilities of the kind discussed is not acceptable, especially as long as it is unclear whether alternative solutions may be available for addressing the most powerful argument supporting segregated facilities: that of reaching otherwise unreachable women.
Is the inferiority test able to deal with the multicultural complexities of this type of case?Given the effect-centred character of this test (that it focuses on the actual (presumed or expected) results of segregation) religiously or culturally inspired motives are not given a specific place in the analysis.Yet, as we can infer from the above, they do figure in the assessment of whether women's power and autonomy are enhanced, and whether 'othering' is reinforced.Besides, the first element of the second part of the inferiority test provides room for taking into account the often vulnerable economic and social position of ethnic minorities.

Overall conclusions
In the preceding sections we have considered the acceptability of three different state-supported sex-segregation practices.For this purpose we used what we call the inferiority test as a specification of equality and non-discrimination standards.This test is inspired by the anti-domination or anti-subordination approach in critical legal studies and US SC case law.
The aim of applying the inferiority test was to assess the legal acceptability of sex-segregation practices according to the same standard, irrespective of whether these practices are quite common in the dominant culture or more recently introduced (again) by or because of minority cultural groups.It turns out that the acceptability of sex segregation in football, primary and secondary education and integration courses needs a carefully balanced appraisal.This has to do, firstly, with the variety of possible or actual aspects of each practice and, secondly, with the multi-layered inferiority test.The result is that the application of the inferiority test to these three cases reveals that in general facilities for boys or men only will not pass the test, as it will probably perpetuate their dominant position.The inferiority test does not offer a clear-cut answer as to whether provisions for girls or women only are acceptable or not, although the answer tends to be negative.
In these conclusions we start by comparing the application of the inferiority test to the three sex-segregated practices, in order to see whether differences and similarities between the cases exist that could explain the applicability and outcome of the inferiority test.We will conclude with a more general analysis of whether the inferiority test is useful as a standard to assess the legal acceptability of old and new sex-segregated practices in a multicultural society.

Comparison of the application of the inferiority test
We start with a comparison of the three cases of sex segregation: what are the differences and similarities in the discussions on sex segregation in football, education and integration courses?

Comparison of the arguments
The first striking factor is that the football case reveals that there is no principled opposition to all-female football teams as such, while -in contrast -mixed football is contested.This factor confirms the rather uncontested situation of female and male football teams.It differentiates sexsegregated sports from the other segregation practices in education and integration courses, which take mixed practices as a starting point and (additional) single-sex provisions as the exception.
However, there is also a similar starting point in the football and integration course cases, which does not return in the education case.This is the appreciation of the participation of girls and women as such, if necessary in a single-sex environment, rather than that they do not participate at all.Underlying this accessibility argument is the idea that sport and integration courses are valuable and important.The difference between these cases is that single-sex football teams are socially and culturally more readily accepted (as a valuable addition) for some age groups, compared to all-women civic integration classes (as an objectionable concession) for some minority groups.
In education the argument of accessibility does not seem to play a role (anymore).Although the practice of education is without doubt of value, there is obviously the assumption that non-participation is not an alternative, which is self-evident in a situation of mandatory education in a school setting, but not when there is the option of home education. 97esides the accessibility argument and the argument of choice between participation and non-participation, other arguments appear in more than one case, although sometimes in a different way.An obvious argument in favour of sex segregation that plays a more or less prominent role in all the case studies is the undesirable physical closeness of the sexes in mixed settings.This could be considered undesirable for reasons of morality or decency (in the case of sport most clearly, to a lesser extent in education, and probably also in integration courses), because of negative effects on achievements (in the case of sport and education, but maybe also in the case of civic integration courses) and/or motivated by religion or culture (in all cases).Especially (average) physical differences (in strength, speed and stamina) are highlighted in sexsegregated sports, but also in sex-segregated education (because of differences in mental development).In both cases those (average) physical differences are related to fair opportunity or competition.
In the case of education, however, the correction of undesirable socialisation plays a more prominent role than physical differences in arguments for (and against) sex segregation.Just think of boys considered to be victims of feminization, and girls thought to be victims of male dominance, especially in 'male' subjects like ICT, maths et cetera.And think of the ethnic or religious minority boys and girls who are considered to be victims of the majority culture in mixed schools.The correction of undesirable socialisation also plays a role in football (to combat stereotypical assumptions about the other sex) and in civic integration courses (to underline gender equality), but in those cases the correction is primarily sought in mixed provisions instead of sex-segregated practices.
Closely related to the correction of undesirable socialisation is the argument of choice, mostly put forward by the proponents of single-sex practices.However, critics tend to wonder whether there really is a choice.
Just as choice figures in arguments in all for and against sex-segregated practices, this is the case as well with equality arguments.This makes the application of the inferiority test significant, for this test is in fact an elaborate equality standard, in which choice plays a fundamental role in the second part of the test.We will continue with a comparison of the application of this test on the different cases.

Comparison of the application of the test
When applying the two-pronged inferiority test to the three cases of sex segregation, the first part of the test -whether the segregated facilities are genuinely equal in terms of all relevant tangible factors -turns out to be problematic in the case of sex-segregated education in the US (due to insufficient governmental supervision) and sex-segregated football (which seems to be partly due to the relatively small number of female players).
The second part of the test -whether segregated facilities create or perpetuate the subordinate position of women instead of diminishing their dominance by men -concerns three clusters of factors.
The first refers to the effect on the social and economic position of women.Sex-segregated football practice does not seem very relevant at the level of amateurs, although negative effects cannot entirely be excluded, for instance because amateur sport may function as a breeding ground for professional sports practice.In the case of sex-segregated education it depends on the effects: when it diminishes the subordinate social and economic position of women it is acceptable, whereas sex-segregated education with the opposite impact is not justified.In the case of sex-segregated integration courses the social and economic position will be improved, assuming that these courses are effective and that participation already diminishes the subordinate position of women.Mixed integration courses could even be better, as is co-education, provided that the teaching is gender inclusive, in order to offer social and economic equal opportunities.
The second factor refers to power and autonomy.In all three sex-segregated practices, the autonomy of women is enhanced when she is able to choose for an all-women facility as a real additional (and substantially tangible) option to mixed practices.This could be even more positive when the sex-segregated facility is meant and suitable to empower women.When these conditions are fulfilled, integration courses for women only could be acceptable according to this factor of the inferiority test.An all-girls school that confirms the subordinate position of girls is not acceptable according to this factor.
Finally, the third factor concerns 'othering'.In all three sex-segregated cases there seems to occur gender stereotyping and/or gender role confinement in such a way that it affirms the hierarchy between men and women.This varies from modesty rules only for girls and women to the opinion that girls are poorer footballers or maths students and victims of male and cultural oppression.Also, male football teams or education for men only will most likely not pass the test of othering, as it will maintain the privileged and dominant position of men in relation to women.In this sense sex segregation seems to confirm othering, although sex segregation is not always inspired by the idea that women are less worthy than men.In contrast, sex segregation is sometimes promoted (as a temporary measure) to overcome othering, in order to reach gender equality in the future.This type of sex segregation seems to undermine othering, although it uses gender generalizations.

Usefulness of the inferiority test
In evaluating the usefulness of the inferiority test on the basis of applying it to the three cases above we would like to point out the following.
To start with, the multilayered character of the inferiority test seems to be quite useful in providing an analytical tool to assess the merits of the multifaceted arguments for and against sex segregation which are often involved.This also seems helpful in avoiding prejudicial assessments of new and often contentious practices of sex segregation as compared to familiar and uncontested ones.
Secondly, given the impact-oriented character of the test, in quite some respects the assessment of whether its criteria are met depends to some or even to a large degree on the availability of empirical data (for instance whether sex-segregated education promotes the selfesteem of girls).Where such data are not available or point to different conclusions this assessment has to be made on the basis of the 'presumed' or 'expected' effects of the segregated practice at stake.In this respect judicial rules regarding burdens of proof and other procedural rules will have to decide how much and which kind of information will suffice.As 'non-segregation' is the point of departure, the initial burden of proof should be on proponents of a segregated practice to show its positive impact.
In the introduction we promised to pay specific attention to the question whether the inferiority test is able to deal with multicultural complexity by ensuring that attention is being paid to minority ethnic, cultural and religious perspectives (cultural for short).The analysis of the case studies shows the inferiority test to be open to such perspectives, though not in every respect.When assessing the impact of segregated practices on the social and economic position of women or their autonomy and power, the specific position of cultural minorities can be given due consideration.Yet, the focus is clearly on the effects of the practice on the position of minority women, not men.This does not seem to be very problematic as long as we realize that the inferiority test tells us whether sex segregation is acceptable from the perspective of sex equality, not whether it could have merits from other perspectives.(See the example of minority boys' schools perhaps being beneficial to improve the social and economic position of minority boys.Even if this would be the case, if the practice is detrimental to the position of minority girls and/or confirms gender hierarchy within minority groups it is not acceptable and we should look for other ways to improve their position.) Arguments for sex segregation inspired by religion or culture also seem to be given adequate space under the inferiority test, especially under the element of 'autonomy'.Here again the focus of the test on women requires women to be the ones supporting sex segregation.Nevertheless, the requirements regarding the avoidance of 'othering' will generally defeat their claims for sex segregation if they are based on traditional views on the proper position of men and women, which are often part and parcel of more orthodox religious views. 98 last, fundamental point we would like to raise regards the question of how to proceed if different elements of the inferiority test point in different directions.If the first element of the test is negative (genuine equality of the tangible factors) while the others are positive, a remedy is rather easily available by upgrading and equalizing all the facilities, financial resources and so on for both sexes (for example, for men's and women's football).
If the three elements of the second part of the test are at stake this is much more problematic.To give an example: can a detrimental effect in terms of supporting notions of sex role confinement (e.g. in segregated civic integration courses) be compensated by a positive effect on (some) women's autonomy or social and economic position?'Non-segregation' being the default position suggests that all the requirements of the inferiority test have to be met in full, but after considering the three cases we would opt for a refinement which is inspired by our case study on sex segregation in education, more specifically the issue of separate maths classes.Though not all elements of the inferiority test are met, the detrimental effects in terms of 'othering' seem pretty limited whereas the potential gains in terms of the other elements could be significant.And in the long run, if segregated maths classes prove effective, improved performance by girls in maths may diminish 'othering'.Besides, alternative non-segregated solutions for improving the performance of girls have been tried rather unsuccessfully for some time, so segregation is not the 'easy way out'.And perhaps in time we will not be in need of separate maths classes for girls at all.
We suggest that the inferiority test should allow for this situation by adapting it slightly: limited detrimental impact can be saved, but only if the following conditions are met: the interest served by the segregated practice is sufficiently substantial to offset a limited negative impact (= proportionality requirement); -alternative, non-segregated solutions are not a real option (= subsidiarity requirement); -the practice is subject to a periodic evaluation to see whether it can be discontinued.
The inferiority test as modified could thus also provide a way out of the dilemmas involved in providing segregated civic integration courses or not.Fully segregated integration courses for men and women still would not pass the test, whereas a more limited and incidental use of segregation (for part of the time and in specific subjects only, like the maths classes for girls) probably would.The detrimental effects in terms of the elements of the inferiority test would be much more limited, whereas the positive effects in terms of making the courses as effective as possible by giving women more space to express themselves can be served at the same time. 122 ) 2008-2009, no.103, pp.8292-8293, <http://ikregeer.nl/document/HAN8469A35>,last accessed on 3 May 2010.92 Regarding the former see Einde aan subsidie gescheiden loketten Utrecht, <http://binnenland.nieuws.nl/568142>,last accessed on 7 January 2010; for the latter see the letter by the Minister of Integration: 'Minister van Wonen, Wijken en Integratie, Brief aan de voorzitter van de Tweede Kamer d.d. 1 juli 2009', Kamerstukken II (Parliamentary Papers) 2008-2009, 31 143, no.63 en 'Brief d.d.17 juli 2009 aan de gemeenten over seksegescheiden inburgering, kenmerk DGW/I&I/2009043114'. 93 The majority of arguments put forward are very well summarized in a publication by E-quality, an NGO working in the field of emancipation, the family and diversity: Handleiding apart integratie-aanbod m/v.Voordelen, nadelen en juridische (on)mogelijheden, December 2009, <http://www.e-quality.nl/e-quality/pagina.asp?pagkey=140380>, last accessed on 7 January 2010.94 Quoted in 'Zoveel mogelijk mensen zo goed mogelijk laten inburgeren', NRC-Handelsblad 14 juli 2009, <http://www.nrc.nl/binnenland/article2300333.ece/Zoveel_mogelijk_mensen_zo_goed_mogelijk_laten_inburgeren>, last accessed on 7 January 2010.
Arguably, the Dutch participation in the Women's UEFA Cup in 2009 provided just such an impulse.59 The Observer (of 25 June 2006) mentions a decline in single-sex state schools in the last 40 years from 2,500 to 400. 60 Rebecca Rogers mentions the exceptionally high number of girls' schools in Ireland in the early 1990s: half of the total number of public and private secondary schools were single-sex schools.R. Rogers, 'Girls' schools', Encyclopedia of Children and Childhood in History and Society, <http://www.faqs.org/childhood/Fa-Gr/Girls-Schools.html>, last accessed on 6 October 2009.In Belgium single-sex schools are still in the majority (The Observer, 25 June 2006).61 Solomon mentions 88 public schools in the US that were either completely single-sex or had single-sex classes for some subjects in 2003-2004.
L. Solomon, 'Trial Run: Schools split Class by Gender', South Florida Sun-Sentinel, 2 September 2003, as cited by N. Levit, 'Embracing Segregation: The Jurisprudence of Choice and Diversity in Race and Sex Separatism in Schools', 2005 University of Illinois Law Review, p. 4, footnote 15.See also Smithers et al. who analyzed reviews from several Anglo-American countries about the evidence of advantages in single-sex and co-education.A. Smithers et al. (eds.),Theparadox of single-sex and co-educational schooling, 2006.128 74 75 Burgin, supra note 65, p. 834.76 Levit, supra note 68, pp.7-10; L. Nappen, 'Why segregated schools for gay students may pass a "separate but equal" analysis but fail other issues and concerns', 2005-2006 William & Mary Journal on Women and Law, pp.101-134, p. 125.77 Blakemore et al., supra note 73, p. 388.78 Smithers et al., supra note 61, p. iii.79 Blakemore et al., supra note 73, pp.370, 376 who write about individual differences in the behaviour of children of both sexes, but at the same time point at general trends of well-behaved girls who do well in school, and boys who misbehave, a situation that is reinforced by the way they are treated by teachers.See also G. Timmermans, cited in the Dutch newspaper Trouw, 24 October 2009.80 Blakemore et al., supra note 73, p. 366, English, supra note 66, p. 2, pp.22-24.
81 Very critical about the claimed success of single-sex education are: P. Haag, 'K-12 Single-sex education: What does the Research say?', Eric Digest, September 2000-EDO-PS-00-9; Smithers et al., supra note 61 and the American Association of University Women, Separated by Sex: A Critical Look at Single-Sex education for Girls, 1998, as discussed by English, supra note 66, p. 22.