EQUAL OPPORTUNITIES COMMISSION v. DIRECTOR OF EDUCATION

HCAL1555/2000

IN THE HIGH COURT OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

COURT OF FIRST INSTANCE

CONSTITUTIONAL AND ADMINISTRATIVE LAW LIST
NO. 1555 OF 2000

-------------------

----------------------

J U D G M E N T

----------------------

INTRODUCTION

1. The Sex Discrimination Ordinance, Chapter 480 ('the Ordinance') was enacted in July 1995. The Ordinance renders unlawful certain kinds of sex discrimination, essentially sex discrimination which occurs in what I will call the 'public aspects' of our lives. The Ordinance does not seek to intrude into family life nor, broadly speaking, into the internal communal life of those who choose by reason of ethnic origins, philosophy or religion, to associate according to particular codes of conduct or belief. The Ordinance seeks to prevent discrimination in those areas of life where individuals may expect to be able to function openly as members of our society on an equal footing with all other members. Inter alia, the Ordinance seeks to prevent sex discrimination in the business of supplying goods, facilities and services to the public (or sections of the public), in the management and disposal of buildings, in the field of employment and in matters of education where that education applies to co-educational establishments.

2. Although the Ordinance talks of discrimination against women, section 6 provides that all provisions of the Ordinance relating to sex discrimination against women shall be read as applying equally to the treatment of men. Similarly, the Ordinance does not restrict its protection to adults. Section 2 of the Ordinance defines men and women as being 'of any age'. Children at school are therefore entitled under the Ordinance to the practical enjoyment of the fundamental right to equal opportunity and treatment without discrimination based on their gender.

3. The Equal Opportunities Commission ('the Commission') is a body formed pursuant to the Ordinance, its statutory mandate being to eliminate sex discrimination in our society and generally to promote equality of opportunity between men and women. It is the Commission which has brought these proceedings for judicial review. It has done so on the assertion that the Director of Education ('the Director') manages a system for the transfer of students from primary to secondary school (called 'the SSPA' system) which discriminates against individual pupils on the basis of their sex and is therefore unlawful under the Ordinance. The Commission asserts that the SSPA system discriminates in the main against girls. However, because of the complex, gender-based construct of the system, it also can (and does) discriminate against individual boys.

4. Under section 5 of the Ordinance, two types of discrimination are rendered unlawful : direct and indirect discrimination. The concept of direct discrimination is defined in section 5(1)(a) :

5. The concept of indirect discrimination is defined in section 5(1)(b) :

Lord Lester, who appeared as leading counsel for the Commission, gave a working definition of indirect discrimination which I can do no better than repeat : "There is indirect sex discrimination where the discriminator treats women and men in the same way (by applying a requirement or condition, such as an age or height requirement, to both of them in the same way) but where the operation of the requirement or condition has a disproportionate adverse impact upon women [or men] and has no objective justification."

6. The enjoyment of equality of treatment free of sex discrimination does not imply that in every instance men and women are entitled to identical treatment. What section 5(1) seeks to prevent is less favourable treatment. The distinction has been recognised by the United Nations Human Rights Committee which, in a body of general comments made by it on non-discrimination (General Comment 18(37), UN DOC CCPR/C/21/Rev.1/Add.1, 1989), said :

The 'Covenant' in question is the International Covenant on Civil and Political Rights ('the ICCPR') to which effect has been given by Article 39 of the Basic Law.

7. The Ordinance is very largely, although not entirely, based on the Sex Discrimination Act 1975 to which I have already made reference. English authorities, dealing with section 1(1)(a) and (b) of the Act, which has identical wording to section 5(1)(a) and (b) of the Ordinance, make it clear that an intent to discriminate is not required to render an act that is discriminatory unlawful. The test is an objective one; causation is the touchstone not intent. In R. v. Birmingham City Council, ex parte Equal Opportunities Commission [1989] 1 AC 1155, Lord Goff, expressing the unanimous opinion of the House of Lord, said (at 1193-1194) :

8. Similarly, in James v. Eastleigh Borough Council [1990] 2 AC 751, Lord Bridge said (at 765/766) :

9. In a report into alleged gender discrimination under the SSPA system published by the Commission in August 1999 ('the Commission's Report'), the following practical observations were made :

10. The test to identify direct sex discrimination has become known as the 'but for' test, that being the test enunciated in Birmingham City Council, ex parte Equal Opportunities Commission (supra) in which - and I repeat - Lord Goff said :

11. In a later judgment in the House of Lords in James v. Eastleigh Council [1990] AC 751 (at 774) Lord Goff expanded upon his earlier dicta :

Lord Goff emphasised that the 'but for' test applied only to direct discrimination, not discrimination alleged to fall under section 5(1)(b) of the Ordinance :

12. The 'but for' test has been approved by our Court of Final Appeal as being the correct test to apply in respect of the Hong Kong legislation as it relates to direct discrimination. See Secretary for Justice and others v. Chan Wah and others [2000] 4 HKC 428 (445 F-G).

13. Article 136 of the Basic Law entrusts to the Government of Hong Kong the formulation of policies on the development and improvement of education. That Article reads (in part) :

14. The Government of Hong Kong, acting through the Director, manages the SSPA system in respect of some 772 primary schools and 433 secondary schools. Although single sex schools are not in any way discouraged, the Director has espoused the principle of co-education and I am told that over 75% of secondary schools participating in the SSPA are co-educational. Mr Pleming, who appeared as leading counsel for the Director, said that the value of a co-educational system of education is not to be undervalued. No argument is taken with that. It follows, of course, that, in promoting co-education, the Director is entitled to take such steps as are reasonably necessary to ensure that the benefits of co-education are maintained and made effective.

15. But that being said, the Government (and therefore the Director) is bound by the provisions of the Ordinance to manage Hong Kong's educational system so that it does not unlawfully discriminate against either boys or girls. Section 21 of the Ordinance provides (subject to certain exceptions which are not relevant to this matter) that it is 'unlawful for the Government to discriminate against a woman in the performance of its functions, or the exercise of its powers'. It follows that the Director and her officers are under a duty imposed by the statute not to discriminate on the basis of gender, either directly or indirectly, by act or by omission, in the discharge of their functions.

16. Sections 25, 26 and 27 of Part IV of the Ordinance, deal specifically with matters of education. While certain exceptions are made for single sex establishments, section 25 states :

The Education Ordinance, Chapter 279, defines all schools as being 'educational establishments' and provides that the Director shall be the 'responsible body' for all schools which are 'entirely maintained and controlled by the Government'. Furthermore, the Director will be treated as having offended the Ordinance if she gives directions to any responsible body for a school to do an act which is unlawful under the Ordinance (section 44) or knowingly aids any such body to do such an act (section 47(1)).

17. Therefore, both generally, as an officer of the Government, and specifically, as the responsible body for all those schools that participate in the SSPA, it is unlawful for the Director to discriminate against individual students (of either sex) in the terms upon which she is prepared to offer them a place in a co-educational secondary school.

18. It was in 1998, when the mechanics of the SSPA system first became widely known, that the Commission began to receive complaints from parents that the SSPA contained discriminatory elements. The complaints were to the effect that the system discriminated against individual students - essentially girls but boys too - on the basis of their sex and was thereby denying those students the opportunity to compete for places in the secondary schools of their choice based solely on their individual abilities, aptitudes and special needs.

19. As a consequence of the complaints of discrimination which the Commission began to receive in 1998, for the first time in its history, it employed the powers given to it under section 70 of the Ordinance to institute a formal investigation. It is to be emphasised that those tasked by the Commission to carry out the investigation were at all times in receipt of the Director's full co-operation. As I have said, the Commission's Report was published in August 1999. This report found (by a majority) that the SSPA system contained three structural elements which discriminated against individual students; the great majority of those students being female. The report found :

20. An executive summary of the Commission's Report concluded :

21. Following the publication of the Commission's Report, there was an active and extended exchange of views between the Commission and the Director. However, although the Director has taken steps to improve what she has called 'gender quality' under the SSPA system, she has remained of the view that there are 'sound and strong education reasons' for supporting the existing system and, if it is found to be apparently discriminatory, for justifying that apparent discrimination. In a letter dated 9 April 2000 addressed to the Commission, the Director said that any such apparent discrimination was justified in order to make 'due allowance for the inherent developmental differences between boys and girls'. Expressed broadly, I understand the Director's position to be as follows :

22. Accordingly, it is the Director's assertion that, while patently there are elements of the SSPA which are gender-based, these elements do no more than remove an initial gender bias in the system and do not, therefore, constitute direct discrimination in terms of section 5(1)(a) of the Ordinance. Alternatively, it is the Director's position that, even if the gender-based elements of the SSPA are directly discriminatory, they are nevertheless justified in terms of section 48 of the Ordinance. This section, which does not appear in the British statute, reads (in part) :

23. The Commission, however, has always taken issue with what has been called the 'late bloomer' effect; namely, that boys mature later than girls. In the Commission's Report, the following was said :

24. As for the director's reliance on section 48 of the Ordinance to justify the gender-based mechanisms in the SSPA, Lord Lester, on behalf of the Commission, has submitted that, as an exception to the fundamental right to equal treatment without sex discrimination, section 48 must be strictly construed so as to avoid the very wrongs which the Ordinance is designed to prevent. It is for the Director to show :

It is Lord Lester's submission that the Director has failed to show that her policy meets those stringent standards. Indeed, it is his submission that the measures adopted by the Director constitute a disproportionate interference with the fundamental right of students to equality of treatment without sex discrimination.

25. At this juncture, what must be emphasised is that the SSPA system is not immutable. It is very much a system in change. Indeed, it appears to be the aim of the Director (in the fullness of time) to so structure the SSPA system that it no longer contains gender-based elements. A recent publication of the Education Commission entitled Learning for life, learning through life recommended abandoning one gender-based aspect of the system by 2005/2006; that is, banding according to gender. In respect of this recommendation, Mr Lee Kwok Sung, an Assistant Director of Education, has commented :

26. This (conditional) prediction recognises that differences in average academic performance between boys and girls, even at or about puberty, are not fixed or genetically predetermined. The Director, of course, does not argue for any simplistic genetic predetermination. Dr Diane Halpern, an educational psychologist at the California State University, has said on behalf of the Director that :

27. In her discussions with the Commission prior to the institution of judicial review proceedings, the Director made it clear that the SSPA was a system in change. Hoping to avoid litigation, the Commission held back in the hope that such changes would assuage its concerns. However, by July 2000 it was the view of the Commission that those aspects of the system which it had earlier determined were discriminatory would remain, at least for the foreseeable future. It appears that the Commission was still receiving complaints from parents. As a result, to use the words of Mrs Anna Wu Hung-yuk, Chairperson of the Commission, the Commission made the decision that it was 'in the public interest, and in the interest of boys and girls alike who may be adversely affected by the system', to seek a declaration from this court that the SSPA did offend the Ordinance in three respects and, if necessary, to seek an order of certiorari to quash those elements of the system that were found to be unlawful.

28. The Commission, however, has not sought a precipitous dismantling of the SSPA, a process that will 'at a stroke' quash central features of the system. Lord Lester, for the Commission, has indicated that this is a case in which declaratory relief may well suffice with liberty to apply. He has qualified this, however, by saying that what the Commission does seek 'at a stroke' is recognition by the Director that direct sex discrimination against girls will be eliminated from the SSPA system within a reasonable time scale, and that, meanwhile, an effective and user-friendly mechanism will be put in place to deal with and to remedy complaints of sex discrimination by individual parents on behalf of their children.

29. The Director has not accepted that any elements of the SSPA offend the Ordinance. Nevertheless, in the event that this Court comes to a different determination, contingency plans have been made. These plans, as I understand them, will attempt, within a reasonable time-frame and with the minimum of disruption, to ensure compliance with this Court's orders and will, in the interim period, deal with and remedy substantiated complaints of gender discrimination made by individual parents on behalf of their children.

THE SSPA SYSTEM : A BRIEF CONSIDERATION OF ITS ORIGINS

30. In Hong Kong today it is mandatory for children to undertake nine years of education, effectively from 6 to 15 years of age. During those years they will transfer from primary to secondary school, invariably from one institution to a new and separate institution. But it has not always been that way. Prior to 1978, only those children in the public school sector who performed to the required standard in a single (and decisive) examination known as the Secondary School Entrance Examination ('the SSEE') were admitted to secondary education. The system of transfer was therefore based entirely on academic merit. Moreover, it appears that academic merit not only indicated which students went on to secondary education but also into which secondary schools those students were admitted. For although all secondary schools in the public sector enjoyed essentially the same funding and the same level of professionalism in staffing, they were not viewed by parents as being equal. Nor rationally, as man-made institutions, despite the best endeavours of the education authorities, could they expect to be seen as being of equal standing. Certain secondary schools gained a reputation for academic excellence. Those schools became the most sought after and each year looked to gathering together an intake of the brightest students in order to cement their reputations. In summary, elite secondary schools emerged and, I am told, still exist today. While an order of merit was never formalised, the education authorities recognised that one did exist by ensuring that certain schools were equipped to foster the brightest while others were better equipped to foster children of less elevated ability.

31. The faults of the transfer system then in existence were well recognised. With the SSEE being the sole determinative, teachers tended to drill their students in the three subjects which were tested in the examination - Chinese, English and Mathematics - to the detriment of a broader primary education. In addition, the building of what had become a semi-formal hierarchy of secondary schools offended more egalitarian principles of education that encouraged secondary schools to take in students of a broader mix of abilities.

32. In the mid-1970s, in anticipation of secondary education becoming mandatory, the authorities looked to a more liberal system of transfer. But a new system could not ignore the realities then in existence. While it was ultimately intended that the meritocratic character of the system should be removed, it was not felt possible to do so entirely at that time. This was reflected in the report of a working party commissioned to propose detailed arrangements for the new transfer mechanism. That report read (in part) as follows :

33. When the SSPA system was introduced in 1978 - some 18 years before the introduction of our domestic sex discrimination laws - it was therefore a system which, because of the constraints of history, was forced to recognise that an elitist hierarchy existed among secondary schools and should (for the time being at least) be maintained. As a consequence, to ensure that the most academically qualified students gained entry to the more elite secondary schools, some form of assessment of academic merit had to be retained in the system. That form of assessment still remains and is fundamental to the SSPA system.

THE SSPA SYSTEM : AN EXAMINATION OF ITS ESSENTIAL ELEMENTS

34. Under the SSPA, Hong Kong has been divided into a number of geographical areas each containing both primary and secondary schools. These areas are known as nets. Parents living in a particular net may only apply for their child to be transferred to a secondary school within their net. The purpose of this restriction has been to shorten travel times between home and school, leaving more time for extra-curricular activities and, where possible, to try and encourage a sense of belonging to a local community.

35. The ability of parents to choose the secondary schools in their net in which (ideally) they would like their child to be educated is, I believe, central to the SSPA system and thereby the issue of discrimination. If parents are entitled to indicate a preference, they are entitled to expect that the system which will attempt to honour that preference will be one that does not discriminate against their child. Nor can it be said that choice is a matter of form rather than substance. Parents are supplied with a list of secondary schools in their net and must list the schools they would like their child to attend in order of preference. I am told that they may list up to 30 schools, an indication that the choice within nets is wide and is therefore an exercise undertaken by parents in the knowledge that it can make a real difference to their child's future.

36. Under the SSPA, priority of choice is decided (in part at least) by academic merit. But students at primary school are no longer subject to the rigours of a single public examination. The SSEE has been replaced by a series of assessments which are devised and marked by the schools which the students attend. These internal assessments (IAs) are three in number. The first assessment is made at the end of the Primary Five year; the second and third during the Primary Six year. To avoid the narrow parameters of the SSEE, students are assessed on all the subjects in their school curriculum with the exception of Physical Education and in some schools, Religious knowledge and Putonghua. The marks are then standardised, different subjects being given different weightings.

37. It is to be stressed that in co-educational primary schools, having been taught together, boys and girls take their assessments together. No segregation exists in the teaching of the students.

38. Mr Lee Kwok Sung has listed the essential characteristics of the IAs. As they are designed by individual schools to test children on the broad range of subjects that they have been taught, they demand the employment of memory as opposed to demonstrations only of abstract verbal or numerical reasoning. Invariably, descriptive answers are required. In short, they appear to follow the model of traditional pen-and-paper tests. However, Mr Raymond Lam, an Assistant Professor in the Department of Education at Hong Kong University, has said that it would not be 'educationally justified' to suggest that the IAs consist only of open-ended questions which involve a great deal of writing. In this regard he has noted :

39. In the mid-1970s, when the SSPA system was in the process of construction, while the public accepted the demise of the SSEE and its replacement by a series of assessments devised by teachers who knew their pupils, there were doubts as to the fairness of the proposed system. Concerns were expressed that some primary schools might mark more leniently in order to boost the merit ratings of their students or that there may be similar internal manipulations. As a result, a centrally-administered test known as an Academic Aptitude Test (AAT) was devised. The purpose of that test was simply to monitor or scale the IA scores given by individual primary schools so that they could be compared fairly with the IA scores given by all other primary schools. In a report dated 5 January 1981, a working party set up to review the SSPA system, explained the purpose of the AAT in the following terms :

40. The working group described the nature of the AATs in the following terms :

41. AATs have a multiple choice format; descriptive answers are not required. They do not test a range of subjects in respect of which students should have been tutored by their schools. To employ the words of Mr Lee Kwok Sung, who also listed the essential characteristics of these tests, they have been designed to emphasise reasoning and to test reasoning ability. As a result, said Mr Lee, although they have never been used for that purpose, certainly not directly, they constitute a better test for predicting academic potential than the IA. By that I take him to mean that, while all tests must assess present ability, because AATs focus essentially on innate powers of reasoning, they are a better diagnostic tool for identifying those students who have not yet realised their full potential but have the innate ability to shine academically if suitably encouraged and nurtured.

42. I have spoken of AATs in the present tense. However, with effect from July 2000, it has no longer been a requirement for students to sit them. But a monitoring device of some kind is still required. The AAT score for each primary school assessed over its final three years is that device. As such - as a body of historical data - AATs continue to be used and, I understand, will be so used for the next few years.

43. In summary, leaving aside gender considerations, the process for assessing academic merit under the SSPA system is as follows :

44. A student's final SSPA score dictates into which band he or she is placed. Originally there were five bands. That has changed and now there are three bands. To explain the allocation : the top third of students are placed into Band One, the second third into Band Two and so on. The bands serve two purposes. They serve to award merit but at the same time they introduce a substantial element of what has been called 'randomization' into the system of allocating secondary school places.

45. Banding recognises merit because all students in Band One are allocated a secondary school place before students in Band Two. It follows that a student - boy or girl - who is placed into Band One will be far better positioned to be given one of his or her top choices of school than a student in Band Two.

46. The purpose of 'randomization' within each band is to attempt to ensure that students of a broader mix of abilities are allocated places in even the most popular schools. "Randomization" works on the basis that there are no merit rankings within each band. The order of preference within each band is determined solely by the random allocation of a computer-generated number.

47. It cannot be denied that the reduction of bands from five to three and the lack of merit ranking within each band has materially increased the 'randomization' factor and reduced the defining impact of academic attainment. But, having said that, while merit does not automatically translate into a fixed place in an order of priority, it is still, in my view, of fundamental importance to the SSPA system. A parent can still quite properly say to his or her child : "Study hard, do well in your school assessments and your chances of getting into one of the best secondary schools in your area will be assured; it may not be your first choice but it will be one of your top choices". In short, it is apparent that, by its structure and essential character, the SSPA system still encourages individual students - of both sexes - to achieve academic merit in order to secure the best possible opportunities for their future.

48. If co-education is to be effective, it is the policy of the Director that gender ratios in co-educational secondary schools should remain balanced. It does not appear that strict gender quotas were always applied. Today, however, as I understand it, the allocation of places in individual secondary schools should reflect the gender profile of the net in which the school is located. Thus, if, after making allowance for students entering single sex schools, the gender profile is 55% boys and 45% girls, the system will attempt to ensure that the intake into local secondary schools will mirror that profile.

49. The Commission has emphasised that the SSPA system is mechanistic; the allocation of places is made by computer. There is no in-built system of appeal. The discretion vested in individual schools is restricted. This is not to say, however, that there is no such discretion. Under the system, secondary schools are given the power to award a certain number of discretionary places. No doubt the factors which dictate the exercise of such a discretion will vary from school to school but one can image that the desire to keep brothers and sisters together may be one of them. The SSPA also recognises that certain primary and secondary schools may have traditional links and, to encourage such links, both 'feeder' and 'nominated school' mechanisms have been put in place. These mechanisms allow parent secondary schools, in addition to the reservation of discretionary places, to reserve a certain number of places for students who come from the primary schools with which they have a recognised attachment.

50. These discretionary mechanisms, while they add a degree of plasticity, constitute exceptions to the central running of the system. They are not therefore of direct relevance to the issues which are central to this judgment. I have used the qualifying word 'direct' because I appreciate, of course, that these discretionary mechanisms are integral to the SSPA and therefore go to its essential nature.

THE USE OF GENDER-BASED MECHANISMS WITHIN THE SSPA SYSTEM

51. When the SSPA system was put into operation, the use of AAT scores to scale IA grades was managed without any reference to gender. However, in the opinion of the Director, early studies of scores obtained by pupils in both the IAs and the AATs revealed a marked gender bias, the IAs favouring girls and the AATs favouring boys. A 1978 report made findings to the following effect :

52. In light of these findings, the decision was made to introduce a gender factor into the scaling mechanism. Initially, girls in co-educational schools had marks deducted from their scaled IA scores while boys in those same schools had marks added to their scores. Patently, at root, no matter how laudable the purpose of the exercise, the intended effect was to boost the final SSPA scores of boys while reducing those of girls. The average margins were not great but they were still material and would have altered the final merit rankings of a significant number of boys and girls. The method of adjustment, however, was found to be unsatisfactory. Mr Lee Kwok Sung has said that the magnitude of adjustment varied from year to year and relied heavily on human judgment, making the process haphazard. Further studies in 1982 indicated that gender bias was still evident. The decision was therefore made to introduce a system which was described succinctly in a contemporary report in the following terms :

53. As a result of that change, under the present SSPA system, boys and girls in co-educational primary schools have their results scaled separately. There is no need to examine the complex mathematical way in which this is done. Suffice to say that what is called a 'gender curve' is used. The intended effect, however, is the same as that of the original system of gender scaling; namely, to boost the final SSPA scores of boys and, in comparison, reduce those of girls.

54. Banding according to gender has always been a part of the SSPA system. Students are ranked according to their final SSPA score and that score will dictate into which one of three bands they are placed. However, boys and girls are ranked separately in this process. It follows, of course, that the band cutting scores will differ for each sex. For example, to get into Band One a girl may require a final SSPA score of 80% while a boy in her same school may only require 75%. Assuming both of them obtain final SSPA scores of 76% and can claim equal academic attainment, the girl will nevertheless fall into Band Two while the boy will be placed into Band One.

55. The imposition of a gender quota for the purpose of maintaining a balanced ratio between boys and girls in co-educational secondary schools has, along with separate gender banding, always been a part of the SSPA. Indeed, the Commission's Report explained the reason for these two mechanisms being gender-based in the following terms :

56. The practical effect of gender quotas is that, once the quota for a particular sex has been filled, any further student of the same sex who has chosen that school will be denied entry and will be allocated another school lower in that student's list of choices. This will be the case even though there are still places available in the school for students of the other sex.

DO THESE GENDER-BASED MECHANISMS DISADVANTAGE INDIVIDUAL STUDENTS?

57. The Commission, in its report of August 1999, came to a finding that, in general terms, there had been less favourable treatment of girls under the SSPA system, since :

58. The Commission's Report also found that, in individual cases, the gender-based mechanisms of the SSPA system had resulted in less favourable treatment of boys. The Commission was therefore of the view that "the SSPA adversely affected both boys and girls in the way they were treated on the ground of their sex."

59. As for the mechanism of separate gender scaling (which sets the machinery for the use of gender into motion), I have already noted that it was designed specifically to neutralise a perceived advantage enjoyed by girls over boys. As such, it was designed to boost the scores of boys at the expense of girls. In such circumstances, I fail to see how it could not disadvantage a significant number of individual girls.

60. Ms Priscilla Chung Ka Tak, the Director of the Gender Division of the Commission, has said :

Ms Chung has made reference to a number of individual girls who have been disadvantaged in the manner she has described.

61. Professor Raymond Lam, an Assistant Professor in the Department of Education of Hong Kong University, was asked by the Commission to consider the matter of gender bias, if any, resulting from separate scaling. In the affirmation which contains his opinions he has said :

62. Professor Lam reached a conclusion with which I am in full agreement :

63. In respect of separate gender banding, Ms Chung has commented :

Ms Chung has supported her comments with the example of one student, a girl, whose final SSPA score of 93.66 placed her in the female Band Four but would have placed her in the male Band Three.

64. Ms Chung made her comments before the bands were reduced from five to three. But, while the reduction in bands has substantially reduced the dangers to individual students inherent in band cutting, the danger still remains. In short, individual students may still be disadvantaged in the same manner as the female student chosen as an example by Ms Chung.

65. In respect of the use of gender quotas, Ms Chung has said :

Again, Ms Chung has been able to cite an example of a student who has been disadvantaged.

66. Of course, the fact that statistics reveal that a number of girls (or boys) have been disadvantaged by the SSPA system does not of itself mean that gender-discrimination must be the sole cause. No doubt there are a broad range of reasons which may go towards explaining the statistics in a more palatable light. I have had a great deal of evidence placed before me, much of it suggesting that there are indeed more palatable reasons. But I am unable to ignore the individual examples of disadvantaged students put before me by the Commission, the evidence being that their disadvantage is either solely or very substantially gender-based. Nor am I able to ignore the compelling conclusion that a system so integrally based on gender difference, with one of its aims being to create an advantage for boys, must throw off a significant number of instances of individual students being disadvantaged for no reason other than their sex.

67. I would add that the Director appears to accept that there must be (and are) instances in which individual boys and girls are unfortunately disadvantaged by the operation of the system. They are counted, it seems, as minor sacrifices for the greater good of all students who participate in the system.

68. In the circumstances, while it would, in light of the complexity of the SSPA system, be impossible to say that each and every case of unequal treatment is based on gender, I can most certainly say that for a significant number of students gender has been (and continues to be) the sole cause of their unequal treatment.

69. But even if other factors may play a role and as a result gender cannot be defined as the sole cause of disadvantage, that does not alter matters. Provided the sex of an individual boy or girl is one of the reasons why that person has received unequal treatment, that, I believe, is sufficient. In this regard, section 4 of the Ordinance provides that :

The SSPA system, while it has attempted to achieve a situation of overall fairness, has sought to do so quite patently on the basis of gender. In terms of section 4, therefore, the reason for the policy is held to be gender. It is that element of gender which I am satisfied has, either wholly or in part, brought about the unequal treatment of a significant number of girls and, to a lesser extent, of boys too.

BUT DO THESE DISADVANTAGES CONSTITUTE DIRECT DISCRIMINATION UNDER SECTION 5(1)(a)?

70. I confess that, on a narrow implementation of the 'but for' test, it seems clear to me that each of the three gender-based mechanisms in the SSPA is discriminatory in terms of section 5(1)(a) of the Ordinance. The Director, however, denies that any of the mechanisms constitute such discrimination. Mr Lee Kwok Sung has explained the general approach of the Director in the following terms :

71. Clearly, the potential of students (of both sexes) to make the best of their secondary education is not only in the interests of those students but also in the wider interests of our society. The Commission does not take issue with this. It has always been sympathetic to the difficulties faced by the Director. But, as the Commission has said in its report :

72. It is the Director's view, on the basis of statistics which she has accepted, that if, by the time they reach the last years of secondary education boys and girls show no real difference in overall academic ability, that per se constitutes evidence that boys do indeed 'catch up'. The problem has been how best to ensure that boys are able to compete with girls 'on a level playing field' in seeking places under the SSPA system in the most academically challenging schools so that, in the fullness of time, they are, in fact, able to 'catch up'. In the opinion of the Director, if girls, because of their generally better scores in the IAs, are able to secure a disproportionate number of places in the best secondary schools that will present boys with an immediate disadvantage, one that cannot necessarily be overcome. Dr Louis Roussos, a psychologist at the University of Illinois who specialises in the analysis of education data, has explained the Director's quandary in the following terms :

73. In the opinion of Dr Roussos, separate scaling by gender does not offer boys any advantage. As he put it (in answer to the suggestion that gender-based scaling is inherently discriminatory) :

74. The Director, of course, does not suggest that the questions asked in the IAs are in themselves biased in favour of girls. On-going attempts are made to ensure that the physical construct of the IAs are made gender neutral. The Commission certainly found no sign of gender bias in the questions. Curiously too, in a set of notes (dated June 1998) directed to the teaching staff of primary schools, the Director said :

75. Despite this limited endorsement of the IAs as a means of testing 'all-round development', Mr Pleming, on behalf of the Director, has submitted that the IAs, as indicators of overall academic aptitude are flawed. They are flawed because, despite endeavours to make the questions gender neutral, the results show that statistically they reveal a gender bias against boys. That being the case, he said, it is pointless to compare final SSPA results - boys against girls - without first attempting to correct that bias. Separate scaling is necessary to correct that bias. Expressed another way, separate scaling does no more than remove a purely temporary advantage enjoyed by girls. The removal of this temporary advantage means no more than that boys and girls are able to compete against each other on equal terms. Separate scaling by gender therefore goes towards removing direct discrimination rather than creating it.

76. As for the use of separate banding and gender quotas, these are employed to ensure that boys and girls are able equally to enjoy the full benefits of co-education. Again, therefore, there is no direct discrimination : the opposite is the case. In the Director's opinion, it could happen, for example, that girls (because their final SSPA scores are better) may start to monopolise the better schools. This could lead to a drift in those school towards extreme gender ratios which will materially dilute the benefits of co-education. There is also the factor of what has been called the 'stereotype threat'. Dr Halpern has spoken of this threat as being one in which :

She continued by saying :

77. In summary, the Director contends that, based on sound educational reasoning, the gender-based mechanisms in the SSPA do not treat one sex less favourably than the other; they seek instead, when viewed overall, to bring about equality of treatment for both boys and girls. There is, therefore, no direct discrimination in terms of section 5(1)(a) of the Ordinance.

78. In my judgment, however, that contention must fail. First, it does not take account of the underlying jurisprudential principle that a fundamental right such as equal treatment free of sex discrimination is a right which attaches itself to the individual. Second, it fails to meet the requirements of section 10 of the Ordinance, a section which directs the Court as to the manner in which the process of identifying discrimination is to be undertaken. I am satisfied that each of the gender-based mechanisms in the SSPA system have been and continue to be the reason why individual students are the subject of direct sex discrimination.

79. In my judgment, if there is a central pillar around which the edifice of Hong Kong's legal system is built, it is respect for the rights and freedoms of the individual. That is manifest in our instruments of constitution, in our adherence to various international conventions and in our domestic law. As Lord Cooke said in a recent House of Lords decision (R. v. Secretary of State for the Home Department, ex parte Daly [2001] UK HL 26) :

80. It is true, of course, that the Director is obliged, in the creation and management of educational policies, to look to the greatest benefit for the largest number of students. But, as the Commission has noted, that does not have to be incompatible with a recognition and protection of the fundamental rights of each individual. It is apparent to me that the Director has looked essentially at what I will call 'group fairness' and, in so doing, has turned a blind eye to the rights of individual boys and girls not to have their school careers (perhaps profoundly) disadvantaged simply on the basis of their sex.

81. In his defence of the Director's policies, Mr Pleming has spoken of the 'purely temporary' advantage in matters academic enjoyed by girls over boys at about the age of puberty. But, of course, he does not talk of all girls being advantaged or of all boys being disadvantaged. Accordingly, in realistic terms, he is only able to talk of boys in the sense of them being a broad group (for example, the top 30% of boys who have achieved better scores in the AATs than girls) or as a gender. In so doing, as I have said, a blind eye is turned to individuals. But, in my opinion, it is exactly to those individuals which our system of law extends protection.

82. I have earlier pointed to the fact that the Ordinance is very largely based on the British Sex Discrimination Act 1975. It is therefore appropriate, I believe, to bear in mind certain principles contained in a White Paper entitled Equality For Women (Cmnd. 5724, September 1974) which influenced the British statute and therefore, if only vicariously, influenced our domestic legislation. Certainly, Bridge LJ, in the Court of Appeal decision of Shields v. E. Coomes (Holdings) Ltd [1978] ICR 1159 (CA), was of the view that the underlying philosophy of the Act was 'well expressed' in paragraph 16 of that White Paper. I believe it is appropriate to cite that paragraph in full :

83. Article 25 of the Basic Law holds that 'all Hong Kong residents shall be equal before the law'. 'All Hong Kong residents', in my view, equates to each and every resident; in short, to each individual. Article 25 of the Basic Law appears in Chapter 111. In respect of the rights guaranteed in that chapter, Li CJ has said, in his judgment in Ng Ka Ling & Others v. Director of Immigration (1999) 2 HKCFAR 4 (at 28/29) :

In the same judgment, Li CJ spoke of the Basic Law - as evidenced by Article 25 - enshrining 'the principle of equality, the antithesis of any discrimination'.

84. Article 25 of the Basic Law is reflected in Article 22 of the Bill of rights, the Bill effectively bringing the provisions of the ICCPR into our domestic law :

85. These guarantees of equality - the antithesis of discrimination - call for a generous and purposive interpretation by our courts. This is to ensure that each person, adult or child, will enjoy the full measure of those guarantees. As Lord Wilberforce said in Minister of Home Affairs & Another v. Collins MacDonald Fisher [1980] AC 319 (PC), constitutional guarantees of human rights call for :

86. It is not disputed that the right to equal treatment free of sex discrimination is in our society a fundamental right; as Lord Lester expressed it, a right of high constitutional importance. As an individual right, it must follow, in my view, that it is a right which cannot be undermined or negatived by broad assumptions or generalisations. What may be true of a group may not be true of a significant number of individuals within that group. The principle is illustrated in the case of Skyrail Oceanic Limited v. Coleman [1981] ICR 864, the headnote of which reads :

In the course of his judgment, Lawton LJ noted that courts, both in the United Kingdom and in the United States, had adjudged that general assumptions based on gender ('stereotyped assumptions', as they are called in the United States) amount to discrimination.

87. Similarly, in Horsey v. Dyfed County Council [1982] ICR 755, Browne-Wilkinson J, in the Employment Appeal Tribunal, said :

88. The Convention on the Elimination of All Forms of Discrimination Against Women ('CEDAW') was extended to Hong Kong in 1996. Article 10 of the Convention makes it plain that stereotyped concepts of both men and women are in themselves, if not discriminatory, at least the wellspring from which discrimination flows. Appropriate measures should therefore be taken to bring about their elimination. In this regard, Article 10 reads :

89. Article 2 of CEDAW obliges all states parties inter alia :

90. The obligation in the Convention to adopt appropriate domestic legislative measures to protect women from discrimination has been met by the coming into force of the Ordinance. To that extent I am satisfied that the words of the Ordinance are to be construed, if they are reasonably capable of bearing such a meaning, as intended to carry out the obligations contained in CEDAW rather than being inconsistent with them. In this regard, see Garland v. British Rail Engineering Ltd [1983] 2 AC 751 (HL), the speech of Lord Diplock (at 771) :

91. That being the case, I am satisfied that the force of section 5(1)(a) of the Ordinance is not to be deflected by broad assumptions, even if statistically well-founded, that categorise women according to stereotypes.

92. Even if such broad assumptions have some general or statistical validity, they still derogate from the rights of the individual. The principle, I believe, was clearly stated in City of Los Angeles, Department of Water and Power v. Manhart (1978) 98 SC 1370 in which the Supreme Court of the United States held that it was discriminatory for the city water and power department to charge women working in the department a higher pension premium than men because the statistics showed that women generally live longer than men and therefore generally draw on their pensions longer. While that may be true, there was nothing to show that the individual women in the department would live longer than their individual male counterparts.

93. What then of the individual students who have been treated unequally under the SSPA, either wholly or in part, on the basis of their gender? Are they to lose the protection afforded to them by law because the Director has chosen to operate a system driven by a broad concept that in Hong Kong girls generally are more academically developed than boys at about the age of puberty?

94. What of the individual girls who have scored, and continue to score, higher marks in their IAs than boys in their class but who, because of gender scaling, nevertheless receive a lower final SSPA score and find themselves materially disadvantaged in their choice of secondary school? Is it to be said that all of those girls are more mature than the boys and that therefore the adverse scaling did no more than make them equal with the boys? Has biology in every instance dictated their better IA scores? Or may it not be the case that a number of those girls, enjoying no developmental advantage of any kind over the boys, have simply applied themselves more diligently to their school work and ensured they were fully prepared for their assessments? If those girls are treated unequally, it does not spring from a personal advantage which must be negatived, it is due solely to the fact that they are girls. But for that fact they would have enjoyed a better choice of secondary school.

95. Is it not also feasible, indeed probable, that among those girls is a girl who is herself (for any number of developmental reasons) a 'late bloomer' but who has conquered her handicap, if it may be called that, by a regime of study? There can be no question of that girl enjoying any temporary or artificial advantage over the boys in her school. The unequal treatment she has received is due to her sex.

96. In determining whether, under section 5(1) of the Ordinance there has been unequal treatment, a comparison is invariably implicit : the treatment given to a person of one sex is to be compared with the treatment given to a person of the opposite sex . But how is that comparison to be made? Section 10 of the Ordinance directs that :

97. Accordingly, in attempting to decide whether there has been direct or indirect discrimination, the court must seek to find the circumstances, if any, which are effectively the same for the persons involved. Or, as it has been said, 'like must be compared with like'. The wording of section 10 is terse, the effect of its meaning not immediately clear. But as I read it, the purpose is to differentiate between purely personal qualities which separate persons of whatever sex and to find the common ground which will determine whether there has been discrimination. As Lawton LJ stated in Noble v. David Gold & Son (Holdings) Ltd [1980] IRLR 252 (at 254) :

98. Clearly, in deciding whether gender is the cause of unequal treatment as opposed to a lack of personal abilities and aptitude, gender itself cannot be one of the 'relevant circumstances'. That is because gender is the very point in issue. Gender must initially, as a matter of logic, be excluded in order to decide whether it may have been the cause of the unequal treatment.

99. This has been the approach adopted by the English courts in reading section 5(3) of the United Kingdom statute which bears the same wording as section 10 of our Ordinance. In Showboat Entertainment Centre Ltd v. Owens [1984] 1 WLR 384, a judgment of the Employment Appeal Tribunal, the issue was one of racial discrimination but the underlying statutory provisions were the same as in the Sex Discrimination Act. The appellant, Showboat Entertainment, had instructed its manager to exclude all black customers from entering its centre. The manager had refused; he had been dismissed and had successfully complained to an industrial tribunal. Browne-Wilkinson J, in delivering the judgment of the court, said :

100. In James v. Eastleigh Council (supra), the plaintiff went with his wife to a leisure centre run by the council. Both were 61 years of age and retired. The wife was admitted free of charge but the plaintiff had to pay an admission fee as the council only provided free admittance, inter alia, to people who had reached state pension age, which in the case of a man was 65 and in that a woman 60. Lord Bridge, in the House of Lords, said :

101. As I understand it, Mr Pleming has argued that, in order properly to compare like with like, it is necessary for this Court to compare the position of boys and girls after their 'relevant circumstances' have been rendered the same for all material purposes by the scaling exercise. Only then, when they are competing on a level playing field, can like be compared with like. Prior to that, their 'relevant circumstances' are materially different. In my judgment, however, the scaling exercise, which purposefully aims to boost the scores of one sex at the expense of another, no matter how laudable its purpose, is palpably gender-based and must, therefore, be excluded. On a proper application of section 10 of the Ordinance, the 'relevant circumstances' which are the same for all boys and girls is that they are all pupils at co-educational primary schools. That, therefore, is the basis upon which it must be determined whether there has or has not been direct discrimination under section 5(1)(a).

102. If, by comparing like with like, the starting point is that the students are all primary school students sitting exactly the same tests, then the 'skewing' of the results of those tests on the basis of gender, when it results in unequal treatment, constitutes direct discrimination.

103. It is apparent that educational specialists in many parts of the world have had to grapple with differences in the development of boys and girls in their more formative years. In many places what I will call 'remedial' measures have been put in hand; that is, measures aimed at encouraging educational advancement of individual students who have not yet revealed their full potential. Elsewhere, what I will call 'institutional' measures have been put in hand; that is, measures that bring about a fundamental change in the system of examinations, grading or admission allocations. Hong Kong is one area where 'institutional' measures have been put in hand. Another area, historically at least, is, or was, Northern Ireland.

104. In respect of Northern Ireland, mention should be made of a particular judgment of first instance which has been much debated before me. It is the judgment of Hutton J in In re the Equal Opportunities Commission and Others (No.1) [1998] N.I. 223. The facts - in their core issues at least - are similar to the present case. The Education Department in Northern Ireland issued a directive that the top 27% of girls and the top 27% of boys in tests set and administered by the Department were to be offered non-fee-paying places in grammar schools. The applicants, the parents of girls, sought a declaration that the Department had discriminated unlawfully against their daughters. They did so on the basis that, under the system, a number of boys had been awarded non-fee-paying places even though some girls had obtained higher test scores. In response, the Department argued that it offered places to identical percentages of both sexes (to quote from the headnote) "because of the difficulty in setting gender free tests and that while it believed that boys and girls at the transfer stage have equal transfer potential, girls in the transfer procedure age group are more verbal and articulate while boys in that age group find spatial relationships easier to understand and that, because it was not always possible to be certain in advance which items in the tests were more likely to favour one particular sex, it was necessary, in order to ensure equality of treatment, to determine the top 27% of boys and girls separately".

105. Hutton J held (to quote from the headnote again) that :

106. Mr Pleming, for the Director, has argued that the case was wrongly decided and that, as circumstances in Hong Kong are very different, I should give it no weight. I accept, of course, that the education system in Northern Ireland, at the time the matter was brought before Hutton J, was different from the education system that we have in Hong Kong. I further accept that the evidence placed before me as to the statistics available to the Hong Kong authorities, the advice of experts and the like is different too. Nevertheless, I have found support in Hutton J's judgment for a number of my findings concerning fundamental principles of law that support our respective sex discrimination legislation. To that extent only, and certainly not on any factual basis or findings as to children's developmental capacities, the judgment has been of persuasive value.

IS DIRECT DISCRIMINATION AGAINST INDIVIDUAL STUDENTS JUSTIFIED IN TERMS OF SECTION 48?

107. As I stated earlier in this judgment, it is the Director's case that, even if it is found that the gender-based mechanisms in the SSPA discriminate against individual students, that discrimination is justified in terms of section 48 of the Ordinance. Section 48 applies inter alia to matters of education. No provision of like kind is to be found in the British statute. There are therefore no authorities from that jurisdiction on point. Nor apparently has the meaning and extent of the section been considered by any court in Hong Kong. Section 48, in full, reads as follows :

108. As I understand it, the Director relies essentially on subsection (a) but accepts that there may, in part, also be some reliance on (b) and (c).

109. On an ordinary reading of section 48, it is clear that the Legislature has accepted that a restriction on the fundamental right of equality of treatment free of sex discrimination may be lawful but only if that restriction is 'reasonably intended' to bring about what, in the various circumstances contemplated in the section, I will call equality of opportunity. Clearly, it cannot have been the intention of the Legislature that any such restriction should undermine the very purpose of the Ordinance; namely, the prevention of discrimination. As to the true purpose of section 48, Lord Lester has said that it may be found in article 4.1 of CEDAW. I agree with that proposition. As I have said earlier in this judgment, the obligation in article 2 of the Convention to adopt appropriate legislative measures has been honoured by the coming into force of the Ordinance. That being the case, on the principle stated in Garland v. British Rail Engineering (supra), the words in the Ordinance are to be construed, if they are reasonably capable of bearing such a meaning, as intended to carry out Hong Kong's obligations under CEDAW and not to be inconsistent with those obligations. What then does article 4.1 of the Convention state? It reads as follows :

110. The article contemplates that states parties may wish, for the purpose of accelerating de facto equality, to take special measures to bring about such equality. The measures, however, are to be temporary. But the essential qualification is that such measures shall in no way entail as a consequence the maintenance of unequal standards; in short, as I read it, any special measures that are taken must not be allowed to undermine the purpose of the Convention itself; namely, the abolition of sex discrimination. The Ordinance does not speak directly of temporary measures. I believe, however, that such a meaning can readily be inferred. I say so because, in my view, if settled regimes that are discriminatory are to be allowed under section 48, that will result in an indefinite maintenance of unequal treatment and will thereby, in all practical terms, undermine the central purpose of the legislation.

111. I have spoken of 'settled regimes that are discriminatory' falling outside of the ambit of section 48. Can the SSPA (at least in its gender-based mechanisms) be described in that manner? It has been in operation since 1978. As I understand it, banding according to gender and gender quotas have always been a part of the system. Separate gender scaling was introduced as long ago as the early 1980s. While 'randomization' factors have substantially reduced some of the discriminatory sting of the three gender-based mechanisms, all three still remain integral to the system. The system is in the process of change. But those three mechanisms are, it seems, to remain for the foreseeable future. In light of those factors, I believe that the SSPA system as presently constituted may accurately be described as a settled regime that is discriminatory in nature. The consequence of the existence of its gender-based mechanisms is the continued and indefinite imposition of directly discriminatory practices on the boys and girls who are subject to it. On that broad basis alone, I am satisfied that the Director cannot avail herself of the protection of section 48.

112. But if I am wrong in that regard and section 48 falls to be considered without such close reference to CEDAW, it becomes necessary to proceed to define how the section should be interpreted; in respect, that is, of administrative policies conducted under it. As to interpretation, I should say that there has been a large measure of agreement between counsel. That has enabled me to move to what I believe are the essential principles more directly than would otherwise have been the case.

113. The Ordinance, although domestic legislation, is, together with the Basic Law and the Bill of Rights, a repository of fundamental rights that are recognised by free and open societies generally. That being so, guidance in interpretation can be obtained not only from national courts in other jurisdictions but also from supra-national tribunals such as the European Court of Human Rights. However, it is not to be forgotten that such decisions, while they may be of valuable assistance in articulating general principles, are limited in their value in interpreting domestic legislation. In this regard, the following was said by Lord Woolf in Attorney-General of Hong Kong v. Lee Kwong-kut [1993] AC 951 (Privy Council) at 966 :

114. In interpreting a statute such as the Ordinance, which is a repository of fundamental rights attaching to the individual, it is a founding principle that any restrictions on those guaranteed rights must be narrowly interpreted. See Ming Pao Newspapers Ltd v. Attorney-General of Hong Kong [1996] AC 907 (Privy Council) at 917 B-C.

115. When a legislative provision which restricts fundamental rights is used, the burden rests on the person employing that provision to justify its use. In the present case, it is therefore for the Director to justify her use of the gender-based mechanisms which are employed by her in the SSPA system. In so doing, she must provide 'cogent and persuasive' reasons. See R. v. Sin Yau-ming [1992] 1 HKCLR 127 at 145. The European Court of Human Rights has echoed this by saying that 'convincing and weighty reasons' are required to justify any such restriction. See Smith and Grady v. United Kingdom (2000) 29 EHRR 493 at 536.

116. Absent any specific wording in a restriction provision, 'convincing and weighty' reasons must be given to demonstrate; first, that any restriction is demonstrably necessary; second, that it is rational in the sense that it is not arbitrary, unfair or based on irrational considerations, and, third, that it is no more than is necessary to accomplish the legitimate objective, in other words that it is a proportionate response. These principles were stated by Bokhary JA (as he then was) in Association of Expatriate Civil Servants of Hong Kong v. Secretary for the Civil Service (1996) 6 HKPLR 333 (at 351/352) when, in speaking of rights guaranteed by the Bill of Rights, he said :

117. The test of proportionality involves a balancing exercise which means that the court no longer reviews the decision of an administrator, such as the Director, in 'Wednesbury' terms. While therefore the court is not in any way vested with a full power to redetermine issues on the basis that such an assumption of jurisdiction would be 'profoundly undemocratic' (see R (Alconbury Ltd) v. Secretary of State for the Environment [2001] 2 WLR 1389), nevertheless, while full deference is given to the legislature and to the makers of administrative policies, it is for the court to reach its own judgment on whether there has been a breach of fundamental rights. Where those rights are of special importance (and I include freedom from sex discrimination in that category), a high degree of protection is appropriate and the court will give less deference than otherwise would be the case to the primary decision-maker. In this regard, see, for example, the dicta of Lord Hope in R. v. Director of Public Prosecutions, ex parte Kebilene and Others [2000] 2 AC 326 (at 381) :

118. The test of proportionality brings, in the need for a balancing exercise, its own judicial dimensions. These were cogently described by Chaskalson P in the Constitutional Court of South Africa in State v. Makwanyane [1995] 1 LRC 269 (at 316/317) :

119. In his judgment (at 318) Chaskalson P cited with approval the dicta of Dickson CJ, Chief Justice of Canada, in R. v. Oakes [1987] LRC (Const) 477 (at 500) :

120. The test of proportionality therefore demands that the restriction should impair as little as possible the fundamental right or freedom in question. In this regard, La Foret J in another Canadian authority, Tetreault-Gadoury v. Canada (Employment and Immigration Commission) [1991] 2 SCR 22, 81 DLR (4th) 358 (at 372), has said :

121. In Human Rights Law and Practice (1999), referred to by Lord Hope in R.v. Director of Public Prosecutions, ex parte Kebilene (supra), the editors have succinctly summarised the principles of proportionality. I would, with respect, adopt that summary :

122. The phrase 'reasonably justifiable in a free and democratic society' or words to the same effect are often used in constitutional instruments protecting human rights or are used as a yardstick to measure the degree to which those rights may be legitimately restricted. Such phrases are aspirational and apply in all respects to Hong Kong.

123. Section 48, while it permits a restriction on the fundamental right of freedom from sex discrimination, only allows such a restriction if it is 'reasonably intended' to bring about a position of equality. A quality of reasonableness must therefore, in terms, of the section be satisfied by any decision-maker; in this case the Director. Words similar to 'reasonably intended' are found as qualifying words in many legislative instruments that recognise and protect human rights. How then are such words to be interpreted? In Elloy de Freitas v. Permanent Secretary of Ministry of Agriculture, Fisheries, Lands and Housing [1999] 1 AC 69 (at 80), giving the judgment of the Privy Council, Lord Clyde adopted the analysis of Gubbay CJ in two judgments of the Zimbabwe Supreme Court :

124. That threefold analysis been approved and adopted in a recent decision of the House of Lords, R. v. A. [2001] UKHL 25 (at page 4 of the judgment), in which Lord Slynn said :

125. The analysis originally made by Gubbay CJ comes from a distillation of South African, Canadian and Zimbabwean authorities. It has been adopted by the Privy Council and the House of Lords. It appears entirely complementary to the analyses earlier made in our own courts. Accordingly, in interpreting the phrase 'reasonably intended' as it appears in section 48, I shall adopt that same process of analysis.

126. During the course of submissions, there were placed before me a number of highly detailed and academically impressive opinions both in defence of the gender-based mechanics employed in the SSPA and against it. As I have earlier indicated, it is quite apparent that in the field of education internationally the problem presented by developmental differences of boys and girls has been and continues to be a most taxing one. It has never been suggested that the Director has available to her a single and simple remedy that will ensure fairness while doing away with discrimination. But the fact that the remedy is not easy is not (in itself) an answer to the Commission's challenge. Fundamental human rights are at issue, rights of individual boys and girls which are demanding of this Court's protection. In short, administrative difficulty cannot be accepted as a reason under section 48 (or any other provision) of the Ordinance for the entrenchment of a discriminatory regime.

127. In considering the many facets of the mechanisms under attack and the consequences their use has on a significant number of individual students, I confess that I have been troubled by a number of matters. After deliberation, my concern over these matters has led me to the conclusion that the Director has not been able to show me, with convincing and weighty reasons, that her use of the mechanisms meets either the second or third limb of Gubbay CJ's analysis. The areas of concern which have led me to this conclusion include the following.

128. Statistics produced to the Commission by the Director (and published in the Commission's Report) have shown that in the period from 1993 to 1998, girls have, as expected, performed better on average in the IAs than boys. However, rather than yielding the field to boys in the centrally-administered AATs, the figures show that on average they have done better than the boys here too; that is, up to about the 70th percentile. It is only above this percentile that boys have been shown on average to perform better than girls. To quote (in part) from the published statistics, the AAT scores show that :

129. These statistics, in my view, raise the valid question : who, in fact, is being advantaged by the separate gender scaling system? Boys generally, it appears, are not showing, in areas of spatial reasoning and the like, an innate aptitude that sets them on equal terms with girls. Only the top 30th percentile are doing so. On this basis, Lord Lester commented that the system, in practice, serves only a minority of boys, the 'male elite', he called them. There is, in my opinion, force in that comment. Put bluntly, a system which has, in the critically important area of assessing scores, institutionalised measures to the potential (and actual disadvantage of all girls, is in reality, only serving the interests of the top 30th percentile of boys. To me, that appears entirely disproportionate.

130. One of the criticisms of separate gender scaling made by the Commission is that, while it may protect a male elite, it does not, in practice, serve to advantage those boys who are supposed to be the essential recipients of that advantage; namely, 'late bloomers'. Ms Priscilla Chung Ka Tak has explained that this rather curious result flows from the manner in which the scaling process is mathematically calculated :

Ms Chung has then gone on to give an example, showing a class (for illustration only) of three boys. The boy who scores the lowest of the three in the IAs but the highest of the three in the AAT, thus showing that he is a supposed 'late bloomer' will nevertheless receive the lowest scaled score of the three and thus be likely to be allocated the least academically challenging secondary school.

131. I appreciate that figures of this kind, while they serve to illustrate a principle, are not true in every instance. I readily accept that there will be boys who are 'later bloomers' who through a fortuitous mix of scores in the IAs and the AAT (now, of course, assessed on a purely historical basis) will be advantaged. But the point, I believe, is well made that the system does not act as comprehensively as the Director would wish to serve those boys who will in due course, given the necessary protection, prove their full worth

132. The Director has placed considerable reliance on the fact that, given the opportunity, on average, boys will be competing on equal terms with girls by the time they come to sit their examinations in the last years of secondary school. In that reliance lies much of the rationale for the use of tools that constitute direct discrimination in the SSPA system. But is there, in fact, clear and convincing evidence of a broad-based 'catch-up' by boys in the secondary years? In this regard, Ms Geraldine McDonald, an educational expert, who has been awarded the Companion New Zealand Order of Merit for her services to education, and is a noted academic, has conducted a detailed analysis and has come to a conclusion that does not support the Director. Upon the basis of her detailed analysis, which I found highly persuasive, she has concluded :

133. In the result, it seems to me that much of the rationale for the introduction of, at least, separate gender scaling is founded upon figures which are not as conclusive as the Director would hope; indeed, they are figures which, upon close analysis, appear to undermine that rationale.

134. At first blush, a fixed gender quota seems benign enough. But I hope I have already made clear the very real disadvantage that it can have. For example, a girl who has, through academic merit, warranted a privileged rank in her choice of school can be denied her choice simply because the gender quota in her school has been filled, this despite the fact that boys in her class with lower scores are free to take up places in that same school. Professor Hau Kit Tai, a professor in the Education Psychology Department of the Chinese University of Hong Kong, has in considerable detail argued the case for the Director that, without fixed gender quotas, there is a real risk of gender ratios swinging to extremes, to the extent even, at the far end of the spectrum, that a co-educational school may become a single sex school. However, with respect, it seems to me to be a rigid theory - for a theory essentially is what it is - which does not give sufficient weight to the natural plasticity of such matters; by that I mean the organic ability for such imbalances, if they threaten, to be corrected. This, of course, can be achieved with the help of measured remedial steps taken by the Director and by the staff of any such threatened schools.

135. As for the beneficial effect of fixed quotas generally, nothing has been placed before me to convince me that they are required in order to ensure that co-education, as a system, works to best effect. The rather sparse survey evidence that was presented tended, if anything, to show that parents would accept an imbalance, a rational one, in order to get their child into a good school.

136. One of the matters that has at all times caused me considerable disquiet is what I perceive to be the 'institutionalised' unfairness of the gender-based mechanics; it is not a case of special remedial measures being introduced, it is rather a case of an entrenched system recognising the need on an indefinite basis to manufacture an advantage in favour of one gender at the intended expense of the other. I have used the word 'intended' because it is apparent to me that on occasions the system works to the disadvantage of those it is meant to favour; namely, boys. To employ military parlance, they become casualties of 'friendly fire'. In summary, in looking at the evidence as a whole, it seems to me that not only is gender unfairness entrenched in the system (and thereby institutionalised within public sector education) but the malaise is compounded by a random element.

137. During the course of submissions a great many matters were raised concerning the respective merits and deficiencies of the SSPA system as presently constituted. I am satisfied, however, that the concerns that I have outlined above are of themselves sufficient to persuade me that the Director has failed to demonstrate that the measures adopted by her which constitute direct discrimination are nevertheless rendered lawful by the provisions of section 48.

138. Are the problems facing the Director intractable? It seems not. The Director herself speaks of changes to the system which will eventually do away with all of its gender-based elements. Elsewhere in the world the same problems have been faced and solutions have been found that do not, because of developmental differences, perpetuate a system of discrimination. While I accept that the Director is involved in a process of change, that change, as it relates to gender-based elements, appears to be very 'open-ended'.

139. The evidence shows that there are remedial measures that can be taken. I confess to finding it puzzling that, in light of the provisions of the Ordinance, the Director has not earlier taken more active steps to implement such measures.

140. The Director, it appears, is concerned that if all the gender-based elements of the SSPA were abandoned, a claim of indirect discrimination could possibly be made by boys. Mr Pleming, in his written submissions, has (in part) expressed the concern thus :

141. Mr Pleming has given further computations of the possible problem. It may well be, of course, that such a claim could be made. But the Director is now well warned of it. She is in a position to take such steps as she deems proper to be able to show that she has taken all reasonable remedial steps, short of entrenched discrimination, in favour of boys, to ensure that boys are given equal opportunities with girls. I certainly do not see her concern as one which excuses the present use of gender-based mechanisms.

CONCLUSION

142. For the reasons set out in this judgment, I am satisfied that all three gender-based mechanisms challenged by the Commission as being discriminatory are contrary to the Ordinance and are unlawful. I will, therefore, grant the declaratory relief prayed.

143. As I said at the outset of this judgment, the Commission does not seek to dismantle the SSPA 'at a stroke'. It is evident that the Director will take such steps as are necessary to honour the findings of the courts. Provisional arrangements have already been made to ensure that from this time substantiated cases of discrimination will be dealt with in such a way as to remedy the effects of such discrimination.

144. At this time, other than the declaratory relief which I have given, I do not intend to make orders for specific relief. As indicated by counsel, such matters can best be dealt with at a further hearing related specifically to matters of relief and to costs. Liberty to apply is therefore given to the parties.