Secretary, Department of Foreign Affairs and Trade v Styles

Re: THE SECRETARY OF THE DEPARTMENT OF FOREIGN AFFAIRS AND TRADE
And: HELEN STYLES and PHILIP ARTHUR HARRISON
No. G1343 of 1988
FED No. 546
Sex Discrimination - Public Service - Administrative Law
88 ALR 621
23 FCR 251

COURT

IN THE FEDERAL COURT OF AUSTRALIA

NEW SOUTH WALES DISTRICT REGISTRY
GENERAL DIVISION
Bowen C.J.(1), Pincus(2) and Gummow(1) JJ.

CATCHWORDS

Sex Discrimination - indirect discrimination - whether the decision to select the second respondent for an overseas posting in the Department of Foreign Affairs and Trade was unlawful under s. 14 (2) of the Sex Discrimination Act 1984 as constituting an act of sex discrimination within the meaning of s. 5 (2) of that Act - construction of s. 5 (2) - meaning of "requirement or condition" - meaning of "substantially higher proportion" - appropriate denominator or base pool to be used to determine "proportion" - meaning of "reasonable having regard to the circumstances of the case" - observations on the contrast between the Australian legislation and British and U.S. legislation in respect of indirect discrimination.

Public Service - Equal Opportunity Program - s. 22B of the Public Service Act 1922 - meaning of "program" in sub-s. 22B (1).

Administrative Law - meaning of "to have regard to" one or more particular matters.

Public Service Act 1922

Administrative Decisions (Judicial Review) Act 1977

Sex Discrimination Act 1984

Civil Rights Act 1964 (U.S.)

Sex Discrimination Act 1975 (U.K.)

Race Relations Act 1976 (U.K.)

Cole v Whitfield (1988) 165 CLR 360

Bath v Alston Holdings Pty. Ltd. (1988) 165 CLR 41

Queensland Electricity Commission v The Commonwealth (1985) 159 CLR 192

Griggs v Duke Power Co. 401 US 424 (1971)

Wards Cove Packing Co. Inc. v Atonio 57 USLW 4583 (1989)

Federal Commissioner of Taxation v Ranson (1989) 86 ALR 267

Perera v Civil Service Commission (1983) IRLR 166

Meer v London Borough of Tower Hamlets (1988) IRLR 399

Clarke v Eley (IMI) Kynoch Ltd. (1982) IRLR 482

Australian Iron & Steel Pty. Ltd. v Najdovska (1988) 12 NSWLR 587

Watches of Switzerland Ltd. v Savell (1983) IRLR 141

R. v Secretary of State for Education; Ex parte Schaffter (1987) IRLR 53

Price v Civil Service Commission (1978) ICR 27

Pearse v City of Bradford Metropolitan Council (1988) IRLR 379

Director-General of Education v Suttling (1987) 162 CLR 427

The Queen v Toohey; Ex parte Meneling Station Pty. Ltd. (1982) 158 CLR 327

Rathborne v Abel (1964) 38 ALJR 293

Kidd v D.R.G. (U.K.) Ltd. (1985) IRLR 190

Goodwin v Jorgensen (1973) 128 CLR 374

HEARING

SYDNEY

#DATE 28:8:1989

Counsel and Solicitors for Mrs. P. Fleming QC and
the Appellant: Mrs. P. Sharp instructed by the

Australian Government Solicitor

Counsel and Solicitors for J.W. Shaw QC and A.L. McSpedden
the First Respondent: instructed by Messrs. Bartier

Perry & Purcell.

ORDER

The appeal be allowed.

Set aside orders 1, 2, 3 and 5 of orders made 18 October 1988 and order that the proceedings be dismissed.

No order as to the costs of the appeal.

Note: Settlement and entry of orders is dealt with by Order 36 of the Federal Court Rules.

JUDGE1

Introduction

This is an appeal from a decision of a Judge of this Court (Wilcox J.) in proceedings brought under the Administrative Decisions (Judicial Review) Act 1977 ("the ADJR Act") for review of a decision under the Public Service Act 1922 ("the Public Service Act"). His Honour's decision is reported: (1988) 84 ALR 408. Sub-section 33 (3) (b) of the Public Service Act incorporates the provisions of the Sex Discrimination Act 1984 ("the Discrimination Act") in respect of appointments, transfers and promotions, being decisions made under s. 50 of the Public Service Act. (It is unnecessary in the present case to consider whether an error of law may have arisen directly under s. 14 of the Discrimination Act; it is sufficient that an error of law may have arisen under s. 33 (3) (b) of the Public Service Act which incorporates the Discrimination Act.)

The appellant, the Secretary of the Department of Foreign Affairs and Trade, contends that the learned primary Judge erred in holding that the decision under review was unlawful under s. 14 (2) of the Discrimination Act as constituting an act of sex discrimination within the meaning of s. 5 (2) of that Act. The first respondent, by Notice of Contention, raises a distinct question. She contends that the learned primary Judge also should have found in her favour that the decision of the appellant was unlawful because he failed to have regard to the objectives of the Equal Employment Opportunity Program of his Department.
We will deal with the facts, then with the grounds of appeal, and finally with the Notice of Contention. We should add that the second respondent took no part in the hearing of the appeal.
Uniformity of treatment is required, and preference and discrimination stigmatised, by the express terms of various provisions of the Constitution (ss. 51 (ii), 51 (iii), 88, 99, 117) and as a matter of necessary implication by s. 92 (Cole v Whitfield (1988) 165 CLR 360; Bath v Alston Holdings Pty. Ltd. (1988) 165 CLR 411). Further, a federal law may be invalid if it discriminates against the States or their agencies: Queensland Electricity Commission v The Commonwealth (1985) 159 CLR 192. It is to be expected that in the Constitution itself, the concept of discrimination will not be attended by any elaboration. This litigation is concerned with a species of discrimination which has been dealt with not in the Constitution but by detailed legislation of the Parliament. This is where one would expect specificity. However, as will become apparent, the occasion for this litigation has been provided by an unfortunate lack of clarity in the terms of the Discrimination Act. The need for clarity can rarely be greater than with legislation of this nature, affecting as it does large numbers of citizens and corporations (and the executive government itself) in everyday dealings and transactions.
The Facts
On 2 September 1987, an officer of the Department of Foreign Affairs and Trade ("the Department") issued a circular seeking "applications for transfers" to certain overseas postings, including the position of Counsellor (Information) in London, designated as a Journalist Grade A2 position. Paragraph 3 of the circular indicated that the objective in postings decisions "is to select the most suitable and efficient officer available to do a particular job". Paragraph 4 was in the following terms:

"4. In reaching decisions on postings, primacy is necessarily given to meeting the Department's operational requirements. At the same time, the Department is concerned to assist officers develop their own professional capacities to the maximum extent possible, as well as make best use of its available staffing resources. In particular, PACs (Postings Advisory Committees) take into account the following factors (not necessarily in order): . job requirements and classification: preference will be given to officers who apply for positions at their substantive APS classification; . relevant qualifications and specialist expertise, experience at home and overseas, and, where relevant, language skills; . officers' posting preferences; . any special circumstances (e.g. family, medical); and . timing and proposed length of posting."

The following sentence appeared in para. 5:

"While officers should feel free to apply for any of the vacancies, applications will be judged against the duty statement/job description for each position." (emphasis in original)

The attached list of postings vacancies contained the prefatory remark:

"Officers are reminded that preference will be given to officers who apply for positions at their substantive APS classifications."

Twelve members of the Australian Public Service applied for the London position of Counsellor (Information); two of them Grade A2 journalists, and ten Grade A1 journalists. Grade A1 is immediately below Grade A2. The first respondent was among those from Grade A1 who applied.
Although the London position was designated Grade A2, the learned primary Judge found that Grade A1 journalists were regarded as entitled to apply. If successful, a Grade A1 journalist would be directed to act temporarily at the higher grade while in London, and would be entitled to a salary at the Grade A2 rate. In the event, the second respondent (who is of Grade A2 and male) was appointed to the London position, and the first respondent's application was rejected.
The first respondent then wrote a letter to Mr. Murray, First Assistant Secretary, Corporate Management Division in the Department, on 22 October 1987 in which she sought an explanation for the rejection of her application for the London posting. Mr. Murray replied in a letter dated 5 November 1987, in the following terms:

"On the matter of the London posting, I have to say that there is no suggestion that you were not judged competent for the position. As you will have seen in the Corporate Management Bulletin there was a substantial number of applications for the position. Two substantive Journalists A2 applied for the position. Both were judged competent. The Committee reached the conclusion that one was not available because of the requirements in Canberra during Bicentennial Year. The other was recommended. Nor was there any judgement as to where you were listed among the Journalists Grade A1."

The learned primary Judge drew the conclusion from this letter that "no real consideration was given" to the application by the first respondent for the London post. The appellant challenges the correctness of that conclusion, pointing to the evidence contained in Mr. Murray's affidavits sworn 3 May and 12 September 1988, as well as the oral evidence given by him. That evidence indicates that he gave at least some consideration to the application by the first respondent. But the question whether Mr. Murray confined the exercise of his discretion to A2 applicants alone does not conclude the matter. Even if discretion was not confined to A2 applicants, it may still have been exercised in such a way as to leave applicants in non-A2 grades with little or no chance of success.
The critical finding of fact made by the learned primary Judge was that "unless an applicant held an A2 grading that applicant had no chance of selection". The evidence given by Mr. Murray in cross-examination bears out the conclusion that A1 applicants had no chance of selection once an application was received from an A2 candidate:

"Once you identify a journalist, to take this present case, at grade A2 and you identify him or her as being suitable and available really no other applicant has any prospect of getting the job? --- I am not required by the law to go through a merit process for transfer and therefore if I am satisfied that an applicant at the substantial (sic) rank is suitable and available then I may move to transfer that person. So provided one of the applicants for a vacancy such as this London job, is a person on the same substantive grade as the vacancy and provided that he or she is suitable and available, then that is the end of the matter? --- That is right."

The first respondent, in written submissions, agreed with the accuracy of statistics submitted by the appellant as to the numbers of male and female journalists of various grades employed in the Department in October 1987. Those statistics can be presented in the following table, not all of which was before the primary Judge:

GRADE No. MEN WOMEN Equivalent to 20 17 3 Grade A2 and above

Equivalent to 53 42 11 Grade A1 and below

The first respondent did not agree with other statistics as to the number of journalists in the whole of the Australian Public Service which were submitted by the appellant. But those other statistics are not relevant to the present proceedings, on the view we take of the construction of the Discrimination Act.
The Concept of Sex Discrimination
Section 14 (2) of the Discrimination Act provides that:

"It is unlawful for an employer to discriminate against an employee on the ground of the employee's sex, marital status or pregnancy -

(a) in the terms or conditions of employment that the employer affords the employee;

(b) by denying the employee access, or limiting the employee's access, to opportunities for promotion, transfer or training, or to any other benefits associated with employment;

(c) by dismissing the employee; or

(d) by subjecting the employee to any other detriment."

Section 14 applies to Commonwealth employees (s. 9 (5)) who are defined in s. 4 (1) as persons who hold an office or appointment in the Australian Public Service or who are employed in a temporary capacity in a Department.
The appellant does not deny that he and the respondents fall within the terms of these provisions. The issue is whether the appellant discriminated on the ground of sex in making the decision under review.
Section 3 of the Discrimination Act states the objects of that Act. They include the elimination, so far as is possible, of discrimination against persons on the ground of sex in the areas of work and the administration of Commonwealth laws and programs, and the promotion of recognition and acceptance within the community of the principle of the equality of men and women.
The concept of sex discrimination is defined in s. 5 of the Act. Sub-section 5 (1) refers to what is often conveniently referred to as "direct discrimination", that is, practices which are discriminatory on their face. Sub-section 5 (1) provides that:

"5. (1) For the purposes of this Act, a person (in this sub-section referred to as the 'discriminator') discriminates against another person (in this- sub-section referred to as the 'aggrieved person') on the ground of the sex of the aggrieved person if, by reason of -

(a) the sex of the aggrieved person;

(b) a characteristic that appertains generally to persons of the sex of the aggrieved person; or

(c) a characteristic that is generally imputed to persons of the sex of the aggrieved person, the discriminator treats the aggrieved person less favourably than, in circumstances that are the same or are not materially different, the discriminator treats or would treat a person of the opposite sex."

Sub-section 5 (2) refers to "indirect discrimination", that is, practices which are fair in form and intention but discriminatory in impact and outcome. This appeal is concerned with the construction of the latter sub-section.

Sub-section 5 (2) provides that:

"For the purposes of this Act, a person (in this sub-section referred to as the 'discriminator') discriminates against another person (in this sub-section referred to as the 'aggrieved person') on the ground of the sex of the aggrieved person if the discriminator requires the aggrieved person to comply with a requirement or condition -

(a) with which a substantially higher proportion of persons of the opposite sex to the aggrieved person comply or are able to comply;

(b) which is not reasonable having regard to the circumstances of the case; and

(c) with which the aggrieved person does not or is not able to comply."

Sub-section 5 (2) resembles in some respects sub-s. 1 (1) (b) of the Sex Discrimination Act 1975 (U.K.). Many of the cases cited to this Court were decided under the British legislation, which provides that:

"1. (1) A person discriminates against a woman in any circumstances relevant for the purposes of any provision of this Act if -

(a) on the ground of her sex he treats her less favourably than he treats or would treat a man, or

(b) he applies to her a requirement or condition which he applies or would apply equally to a man but -

(i) which is such that the proportion of women who can comply with it is considerably smaller than theproportion of men who can comply with it,

(ii) which he cannot show to be justifiable irrespective of the sex of the person to whom it is applied, and

(iii) which is to her detriment because she cannot comply with it."

There are some important differences between sub-s. 1 (1) (b) of the British Act, and sub-s. 5 (2) of the Australian Act, and accordingly care must be taken in the use of the English decisions in the present proceedings. In the first place, the British Act is directed to discrimination against women only, whereas the Australian Act (as indicated by the statement of objects in s. 3) is concerned with discrimination against either sex. When it deals with the respective proportions of women and men who can comply with the relevant requirement or condition, para. (b) (i) of sub-s. 1 (1) of the British Act does so by referring to all those who can comply and requiring a comparison between the proportion of women and the proportion of men who can comply. In the Australian Act, the one form of words has to encompass a double function i.e. to deal with discrimination against women because a greater proportion of men can comply, and discrimination against men because a greater proportion of women can comply. In Australia, unless there is no substantial difference in proportions, in any given case either men or women will be able to show that they have made out this element in their favour.
Secondly, under the British Act, once the complainant proves that a requirement or condition has a disparate impact on women, the burden shifts to the respondent to "justify" the practice; under the Australian Act, on the other hand, the burden of proof does not shift. Thus, in construing the Australian Act, it would be inappropriate to refer to conduct as being "prima facie discriminatory" upon proof of a requirement or condition which satisfies paras. (a) and (c) of sub-s. 5 (2), before one inquires into para. (b). On the contrary, all the elements of sub-s. 5 (2) must be present before it can be concluded that conduct amounts to sex discrimination.
(A further contrast is provided by the United States legislation embodied in Title VII of the Civil Rights Act 1964, 42 USC & 2''2000e-2(a). This makes it an unfair employment practice for an employer to discriminate against any individual with respect to hiring or the terms and condition of employment because of such individual's race, colour, religion, sex, or national origin; or to limit, segregate or classify his employees in ways that would adversely affect any employee because of any of these characteristics. In Griggs v Duke Power Co. 401 US 424 at 431 (1971), Title VII was construed so as to forbid "not only overt discrimination but also practices that are fair in form but discriminatory in practice". This indirect discrimination is said to involve "disparate impact". There remains vigorous disagreement in the Supreme Court as to the placement of the initial burden of proving disparate impact and as to whether the burden shifts to the defendant to justify any disparate impact as to the operation of legitimate business practices: Wards Cove Packing Co. Inc. v Atonio 57 USLW 4583 (1989)).

Thirdly, the Australian Act refers in sub-s. 5 (2) (b) to a requirement or condition which is not "reasonable having regard to the circumstances of the case", in contrast to the wording of sub-s. 1 (1) (b) (ii) of the British Act, which refers to a requirement or condition "which (the respondent) cannot show to be justifiable irrespective of the sex of the person to whom it is applied". We return to the content of the "reasonable" test later in these reasons under the heading "Reasonableness".
We now turn to consider each of the four elements of the concept of sex discrimination posited by sub-s. 5 (2) of the Act.
Sub-section 5 (2) contains four elements. The first is that the alleged discriminator "requires the aggrieved person to comply with a requirement or condition". The second is that the requirement or condition be one "with which a substantially higher proportion of persons of the opposite sex to the aggrieved person comply or are able to comply". The third is that the requirement or condition be "not reasonable having regard to the circumstances of the case". The last is that the requirement or condition be one "with which the aggrieved person does not or is not able to comply". The appellant submits that the first respondent has not made out her case as to the first, second and third elements. She fails and the appeal on this branch of the case succeeds if any of the appellant's submissions are made good.
Requirement or condition
As to the first element in the definition in sub-s. 5 (2) the learned primary Judge held that, in the present case, there was a "requirement or condition", within the statutory meaning, in the insistence that prospective appointees be A2 Grade, or be not less than A2 Grade, or (alternatively formulated) that they be of the same substantive grade as the relevant posting. The appellant contends that this is not a "requirement or condition" because it did not operate as an absolute bar to selection. We agree with the learned primary Judge that something falling short of an absolute bar to selection may be a "requirement or condition". "Condition" is a term used in differing technical senses in various areas of the law: Federal Commissioner of Taxation v Ranson (1989) 86 ALR 267 at 271. Here, like the term "requirement", it takes its colour from the particular statutory context in which it appears.
The English Court of Appeal adopted the construction of "requirement or condition" for which the appellant contends in Perera v Civil Service Commission (1983) IRLR 166 and in Meer v London Borough of Tower Hamlets (1988) IRLR 399. Those cases were decided under sub-s. 1 (1) (b) of the Race Relations Act 1976 (U.K.) which also contains the element of "requirement or condition". In Perera v Civil Service Commission (supra) Stephenson L.J. equated "requirement or condition" to an "absolute bar" (at 170) or, less alluringly, to "a must" (at 169). A mere factor, or set of factors, to be taken into account in reaching the decision would fall short of that meaning (at 170).
The first respondent, on the other hand, points to a different though not, in our view, an inconsistent line of authority in the United Kingdom. In Clarke v Eley (IMI) Kynoch Ltd. (1982) IRLR 482, Browne-Wilkinson J. said of the expression "requirement or condition" that "it is not right to give these words a narrow construction" (at 485); (approved by Priestley J.A. in Australian Iron & Steel Pty. Ltd. v Najdovska (1988) 12 NSWLR 587 at 611). Similarly, in Watches of Switzerland Ltd. v Savell (1983) IRLR 141, Waterhouse J. said that "the statutory words should be given a liberal interpretation in order to implement the object of the legislation" (at 146).
In construing these words, it is essential to bear in mind that the concept of indirect discrimination posited by sub-s. 5 (2) is concerned not with form and intention, but with the impact or outcome of certain practices. In seeking to show that there was no requirement or condition in the present case, the appellant places considerable weight on the terms of the departmental circular inviting applications for the London posting which indicated that non-A2 journalists were eligible to apply, although A2 candidates would be preferred. But to look solely at the terms of the departmental circular is to put misplaced reliance on the formal terms of the criteria for selection, which is both contrary to the legislative intent and would tend to encourage evasion of the operation of the statute. Sub-section 5 (2) is concerned with practices which are fair in form but discriminatory in practice. The correct approach is to look at the way in which those formal criteria operate in terms of their impact in practice: cf. Pannick, "Sex Discrimination Law", pp 43-44. A requirement or condition, therefore, means a stipulation which must be satisfied if there is to be a practical (and not merely a theoretical) chance of selection.
In the present case, that construction of sub-s. 5 (2) requires a more precise identification of what was entailed by the "preference" to be given to A2 applicants. The oral evidence of Mr. Murray indicates that, if there were an A2 applicant, he or she would be selected, and that only if there were no A2 applicants would an A1 applicant stand any chance of selection. As it happened, two A2 applicants expressed interest, with the result that the first respondent, by virtue of having only an A1 grading, had no chance of selection. Therefore, in practical effect, and despite the formal terms of the circular, there was a requirement or condition that the applicant be of A2 Grade, or (which amounts to the same thing) that the applicant be at the same substantive level as the position advertised.
In addition, it is clear from the terms of the departmental circular seeking "applications for transfers" that applicants must already be officers of the Department.
"Substantially higher proportion"
The second element in the definition in sub-s. 5 (2) is that the requirement or condition be one "with which a substantially higher proportion of persons of the opposite sex to the aggrieved person comply or are able to comply". This element requires a comparison between two proportions, using "proportion" to indicate in each case the relationship between two numbers expressed as a numerator and denominator. But it is not plainly stated in the sub- section how the integers in those proportions are to be identified and calculated. In both the Australian and the British statutes, the succinctness of the statutory expression has produced some ambiguity. But, as we have indicated, the difficulty with the Australian provision is increased by the double operation of para. (a) of sub-s. 5 (2), so as to deal with discrimination against either sex.
What is clear is that the numerator in each proportion is the number of men or women (respectively) who comply or are able to comply with the requirement or condition. The uncertainty concerns the appropriate denominator, sometimes described as the "base pool" or "base group".
In our view, the correct construction is that the appropriate denominator (or base pool) is the number of people (men and women) who comply or are able to comply with the requirement or condition. That construction operates directly upon the terms of sub-s. 5 (2) (a) in that it makes use of both groups to which sub-s. 5 (2) (a) expressly refers, i.e. men who comply and women who comply, without adopting statistics drawn from any groups to which sub-s. 5 (2) (a) does not expressly refer (such as all men and women employed by the respondent or all men and women in the population at large). Within the universe of discourse created by the terms of the sub-section, one looks at all those who comply or are able to comply and asks whether of that total those of the opposite sex to the aggrieved person may be described as a substantially higher proportion than those of the sex of the aggrieved person. Applying this construction in the present case, the total of those persons equivalent to Grade A2 and above who comply or are able to comply is 20; the proportion of persons of the opposite (male) sex to the aggrieved person who comply or are able to comply is 17/20 (or 85%); and the proportion of persons of the (female) sex of the aggrieved person who comply or are able to comply is 3/20 (or 15%). The former proportion is substantially higher than the latter proportion. It is a clear inference from the evidence that the same is to be said if one looks only to applicants of Grade A2, the grade to which the requirement or condition actually was directed. The learned primary Judge held (84 ALR at 425-426) that para. (a) of sub-s. 5 (2) was satisfied, whether one treated the relevant requirement as being "A2 Grade" or "not less than A2 Grade". As we have indicated, the requirement or condition was that the applicant be "A2 Grade". However, his Honour's approach to the construction of para. (a) is in essence that which we have taken.
The practical sense of this construction is that it holds up for scrutiny all requirements or conditions which have a disparate impact on men and women, and requires that they be shown to be reasonable. It may be that such disparate impact occurs because the status quo already reflects substantially disparate opportunities for men and women. By taking purely "intra-sex" proportions, on the other hand (which we deal with below), the proportions will differ substantially only if the condition causes the disparity to increase further, whatever may be the present position. But as s. 3 of the Discrimination Act indicates, the intention of Parliament was that, so far as possible, discrimination against persons on the ground of sex, in the area of work, be eliminated, not merely prevented. This contemplates that the status quo itself be scrutinised and that any discrimination (whether already existing or prospective) which is unjust be eliminated.
The appellant contends, however, that the literal words of sub-s. 5 (2) (a) require that both the numerator and denominator comprising each proportion be drawn from statistics relating to either sex alone; that is, that each proportion be an "intra-sex" proportion. This was the construction adopted by Priestley J.A. in Australian Iron & Steel Pty. Ltd. v Najdovska (supra at 609-617) with whose reasons Mahoney J.A. generally agreed; cf. the view of Schiemann J. in the English case R. v Secretary of State for Education; Ex parte Schaffter (1987) IRLR 53 at 56. In Najdovska's case, Priestley J.A. held that the appropriate denominators were the numbers of men and women (respectively) within the relevant jurisdiction (i.e. New South Wales). Applying this construction to the present case, on the basis that there are roughly 8 million men and 8 million women in Australia, one obtains very small proportions; the number of men who comply with the requirement (i.e. 17) divided by the total male population of Australia is 0.000002125%; the number of women who comply with the requirement (i.e. 3) divided by the total female population of Australia is 0.000000375%; and the former is not substantially higher than the latter figure.
There is a number of reasons why we reject this construction. In the first place, it introduces into the calculation groups which are not referred to in sub-s. 5 (2) (a), viz. the total male and female populations of Australia. Therefore, this construction contradicts the insistence of its proponents on a literal interpretation of the statute. The same would be true of any attempt to define the base group by some intermediate criterion (i.e. between the class demarcated by the requirement or condition on the one hand, and the total male and female populations of the jurisdiction, on the other) or by reference to a larger group, e.g. all men and women in the world.
To adapt an example suggested by Australian Iron & Steel Pty. Ltd. v Najdovska (supra), if the employer has a "last on, first off" system of retrenchment such that employees hired on or before a certain date are not liable to be retrenched, then the base group is simply the total number of employees hired on or before that date, not some broader group such as all employees in the company. Hence, while we agree with Street C.J. in Najdovska's case (at 596) that the base group is all people "divided or segregated" by the condition (in the sense that the condition segregates those who comply from those who do not comply), we would have thought that in applying that construction to the facts in Najdovska, the base group would have been the company's employees hired on or before 6 February 1981 (i.e. being those who comply with the condition), rather than all employees of the company.
Caution must, of course, be exercised in assessing how a particular requirement or condition will operate in practice in identifying the class to which it refers. For example, in Price v Civil Service Commission (1978) ICR 27 at 32, Phillips J. suggested that, in identifying the class picked out by a requirement that job applicants be under the age of 28, the better approach would be to consider only those men and women under 28 who were qualified for the job, rather than all men and women under 28. (This was approved by the English Court of Appeal in Perera v Civil Service Commission (1983) IRLR 166 at 170; cf. Pearse v City of Bradford Metropolitan Council (1988) IRLR 379 at 381.) But once the class picked out by the requirement or condition is identified, no other groups enter into the calculation.
We turn to the second of the reasons indicated above. This is that the words used in sub-s. 5 (2) (a) do not necessarily require that each proportion be "intra-sex". There is an ambiguity in the expression in sub-s. 5 (2) (a) "proportion of persons of the opposite sex", as to whether the denominator relates to one sex or both. This ambiguity should be resolved by the consideration that we have just discussed, namely that the statute does not refer to any groups other than men and women who comply with the condition.
The ambiguity may be illustrated by the proposition that "the proportion of carpenters who are men is higher than the proportion of carpenters who are women". That proposition would be demonstrated by taking as the denominator the total number of carpenters, both male and female. It might be objected that this is merely a comparison of numbers, not of proportions, and that the proportions have no significance which could not be conveyed by looking at the numbers alone. We do not think that this is a valid objection. If, for example, there were 900 male carpenters and 100 female carpenters, then on a comparison of numbers there is a difference of 800 which is "substantial". But the difference is substantial because the proportion 900/1000 is substantially higher than 100/1000. If, on the other hand, the total number of carpenters were 10,000 of whom 5,400 were men and 4,600 were women, then the difference of 800 would not be substantial because the proportion 5,400/10,000 is not substantially higher than 4,600/10,000. The point of this illustration is that, in the present context, there is no dichotomy between comparing numbers and comparing proportions, because the element of substantial difference requires that the numbers themselves be expressed as a relationship of parts to the whole, i.e. as proportions.
Thirdly, we note that in the Explanatory Memorandum circulated during the debate in the Senate on the Bill for the Discrimination Act, it was stated, when dealing with sub-clause 5 (2):

"Sub-clause (2) defines discrimination on the ground of sex by a discriminator against an aggrieved person to cover requiring the aggrieved person to comply with a requirement or condition which more persons of the opposite sex to the aggrieved person could comply with, which the aggrieved person does not or cannot comply with and which is not reasonable in the circumstances."

In other words, the concern was with a base pool comprising numbers of men and women in the sense we have discussed earlier in these reasons when giving what we consider the correct construction of the legislation, rather than with a base pool or pools as identified by the appellant.
Fourthly, if the construction contended for by the appellant requires calculations based on the male and female populations of the entire jurisdiction, the result is one which is absurd on the face of sub-s. 5 (2) (a). Sub-section 5 (2) (a) envisages that one proportion may be substantially higher than another. But the supposedly literal approach has the effect of rendering both proportions so small as to be incapable of being substantially different from each other. We have given an example earlier in these reasons. In R. v Secretary of State for Education; Ex parte Schaffter (supra at 56) Schiemann J. sought to meet such an objection by the following analysis:

"In most cases, both X and Y will be very small percentages, but since one is comparing X with Y the difference between one percent and two percent is no less significant than the difference between 30% and 60%. In consequence, if the pool of which the percentages are taken includes all humanity it does not matter if the practice under attack has no impact on the vase bulk of humanity. You may land up comparing two small percentages, but looked at in terms of each other rather than in terms of the whole, the difference between them is significant."

That analysis supposes that in order to say that one proportion is substantially higher than another, one must further divide one proportion by the other and reach a quotient which appears substantial. But the ordinary meaning of "substantially higher" in the present context is a reference to the difference between the two proportions (obtained by a process of subtraction) not a quotient obtained by some further process of division.
Finally, it must be borne in mind that sub-s. 5 (2), dealing with indirect discrimination, is not an exhaustive treatment of the concept of sex discrimination. Sub-section 5 (1) deals with direct discrimination. We have set out the text earlier in these reasons. The provision is expressed more broadly than the relevant provision in s. 1 (1) (a) of the British Sex Discrimination Act 1975. In the British legislation, direct discrimination against a woman is confined to discrimination "on the ground of her sex", whereas the grounds mentioned in the Australian provision in sub-s. 5 (1) go beyond the sex of the aggrieved person to include a characteristic that appertains generally to, or is generally imputed to, persons of the sex of the aggrieved person. A height or weight requirement, for example, may fall under sub-s. 5 (1), as well as perhaps under sub-s. 5 (2), but only for the purposes of the former sub-section would it be relevant to inquire into the prevalence within the total male and female populations respectively of the particular requirement. As we have indicated, sub-s. 5 (2) does not require any statistical analysis beyond the class identified by compliance with the impugned requirement or condition.
(The United States law as to indirect discrimination is spelled out by judicial glosses upon the statute rather than in specific terms of the statute itself; thus it is an unsafe guide when construing the Discrimination Act. However, one may note that under the United States decisions, ordinarily the base group is not the general population but the "qualified" population in the "relevant labour market"; but there is no clear agreement as to what is meant by "qualified" and "relevant labour market": Wards Cove Packing Co. Inc. v Atonio 57 USLW 4583 (1989) at 4585-4587, 4592-4593.)
We recapitulate. Paragraph (a) of sub-s. 5 (2) may invite a variety of constructions. One cause of disagreement is the use of the word "proportion" instead of "number", which may be thought to require a search for groups other than the groups of men and women who comply with the requirement or condition so that a denominator can be found for an "intra-sex" proportion. In our view, the use in the statute of the word "number" would have foreclosed this debate, but, as we have indicated, the word "proportion" is also apt to reach the same result, because in assessing whether the difference in numbers is substantial, those numbers must be expressed as parts of the whole.

The constructions of para. (a) of sub-s. 5 (2) which are based on "intra-sex" proportions would require substantial rewriting of the section in order to indicate what groups are to be used as the denominators. The groups to which the provision refers are those of the opposite sex to the aggrieved person who comply with the requirement or condition, and those of the sex of the aggrieved person who comply. If the proportions are to be within either sex alone, then statutory guidance is presently lacking as to what other groups are relevant. There are many possibilities available to one with the task of redrafting the statute; e.g. men and women employed in similar work (either by the person who stipulates the condition or generally); all men and women employed by the person who stipulates the condition; all men and women who are employed somewhere in the jurisdiction; or even all men and women in the jurisdiction. The statutory silence on this question cannot readily be inferred as an invitation to judicial solution. The supposed distinction between comparing proportions and comparing numbers would be a slender basis for such a construction.
Further, as we have indicated, the construction which we accept more readily advances the relevant objects of the Discrimination Act, as spelled out in s. 3. The appeal fails, as regards the second element in sub-s. 5 (2) of the Discrimination Act.
Reasonableness
The third element in the definition of sex discrimination in sub-s. 5 (2) is that the requirement or condition be "not reasonable having regard to the circumstances of the case". As Wilcox J. held (84 ALR at 429) the test of reasonableness is less demanding than one of necessity, but more demanding than a test of convenience. We agree. The criterion is an objective one, which requires the Court to weigh the nature and extent of the discriminatory effect, on the one hand, against the reasons advanced in favour of the requirement or condition on the other. All the circumstances of the case must be taken into account.
However, in our view, it is unnecessary to decide whether "reasonable" differs from "justifiable" in s. 1 (1) (b) (ii) of the British legislation.
Wilcox J. held in the present case that the requirement or condition that applicants be Grade A2 was not reasonable in the circumstances of the case (84 ALR at 431). His Honour said that the arguments put to support the requirement or condition depended essentially on two factors: expense and tidiness of administration.
The expense was said to arise out of the additional salary payment which would be made if an officer of lower than A2 Grade were appointed to higher duties. But his Honour found that if the Department was to fill its A2 positions, then it would have had to make temporary appointments of lower grade, there being insufficient A2 journalists to fill the available posts (84 ALR 430). Therefore, it would have made no difference in expense overall if the London posting in particular was filled by temporary appointment of an A1 Grade journalist.
In respect of the factor of tidy administration, we agree with the learned primary Judge that such considerations alone would be insufficient to render the requirement or condition reasonable in the circumstances of the present case. The evidence of Mr. Murray was that 12% of all transfers made for the period January to August 1988 were temporary appointments at higher duties. As his Honour said, the situation was already untidy, but manageable.
Nonetheless, in our view, the reasons advanced in favour of the impugned requirement or condition go beyond expense and tidy administration. In his affidavit sworn 3 May 1988, Mr. Murray said that:

"Officers are, in my opinion, entitled to expect the Department to find work for them at their substantive level, subject to certain limitations relating to incapacity, redundancy and their being unattached officers of the Service as, for example, under Part IV of the Act. However, the Department has no responsibility to employ an officer in a position at a higher level than their substantive level. Officers are required temporarily to perform the duties of a position higher than their substantive position only when directed by an officer authorised to give such a direction. They then temporarily perform the duties, as a matter of duty, in order to meet operational needs of the Department."

The practice of appointing officers at their substantive grade is certainly conducive to tidy administration, but it also conforms to a precept of fairness that persons be employed according to the substantive level of their qualification. The force of the passage from Mr. Murray's affidavit which we have extracted is that one of the rewards of a higher qualification is to be appointed to a position for which that qualification is demanded in preference to a person who, though perhaps having the capacity to fulfil the position, lacks the higher qualification. His Honour said that "A tidy mind will readily see the virtue of matching like with like" (84 ALR at 430). But that is a virtue which would appeal also to a fair minded administrator. The fact that the practice was not invariable does not diminish the importance of the practice as a general rule of fairness, even if it diminishes the tidiness of the administration.
It should be emphasised that the requirement or condition in the present case is one based on merit, rather than, for example, purely on seniority. The entitlements to which Mr. Murray referred derive from the meritorious use of effort and ability. Different considerations might apply where the requirement or condition is not related to merit.
In speaking of reasonableness, unlike the previous elements of sub-s. 5 (2), one is not necessarily trying to assess states of affairs and their consequences. The task is rather to assess what reasonableness requires of particular people, which depends essentially on what responsibilities they have: cf. Finnis, "Natural Law and Natural Rights", 175. It is thus not necessary to consider the extent of any staff resentment or industrial disharmony which might arise from alternative selection policies, although any evidence of this would further strengthen the appellant's case: cf. Director-General of Education v Suttling (1987) 162 CLR 427 at 433.
In our view, the precept of fairness to which we have referred, when weighed against the discriminatory impact, is sufficient to render the requirement or condition reasonable in the circumstances of this case. The third element in sub-s. 5 (2) is thus not satisfied. On this ground, the appellant succeeds.
Inability of the aggrieved person to comply
The fourth element in sub-s. 5 (2) is that the requirement or condition be one "with which the aggrieved person does not or is not able to comply". The requirement or condition was that applicants be of A2 Grade; the first respondent was not of A2 Grade. Plainly, the fourth element is satisfied, and there was no submission to the contrary.
Conclusion as to Discrimination
It follows that the appellant has succeeded on this branch of the appeal. It remains to consider the issue raised by the first respondent's Notice of Contention.
The Equal Employment Opportunity Program
The first respondent seeks to support a result in her favour upon a ground she failed to make out before Wilcox J. By her Notice of Contention, she contends that the appellant's decision was unlawful by reason of a failure to have regard to the objectives of the Equal Employment Opportunity Program of the Department, and by reason of a failure to use certain forms issued pursuant thereto. She argues that by contravening s.22B (5) of the Public Service Act, the decision to transfer or promote an officer under s. 50 (3) of that Act was tainted by an error of law within the meaning of the ADJR Act.
Sub-section 22B (5) of the Public Service Act provides that:

"The Secretary of a Department shall take any action necessary to give effect to the equal employment opportunity program for the Department and any person who exercises powers in relation to employment matters in the Department shall have regard to the program in exercising those powers."

"Equal employment opportunity program" is defined in s. 22B (1) of that Act as follows:

"equal employment opportunity program", in relation to a Department, means a program designed to ensure that -

(a) appropriate action is taken to eliminate unjustified discrimination against women and persons in designated groups in relation to employment matters in the Department; and

(b) measures are taken to enable women and persons in designated groups to -

(i) compete for promotion and transfer in the Department and in the Service generally; and

(ii) pursue careers in the Department and in the Service generally, as effectively as other persons."

Sub-section 22B (1) further provides that:

"'program' includes -

(a) the particular objectives to be achieved achieved by the program;

(b) the policies to be adopted, and the procedures to be followed, to achieve those objectives;

(c) the quantitative or other indicators against which the effectiveness of the program is to be assessed; and

(d) the allocation of staff and other resources to the task of giving effect to the program."

Sub-section 22B (1) also contains a definition of "employment matters" which includes, inter alia:

"(b) the promotion and transfer of -

(i) officers to offices in the Department; and

(ii) officers in the Department to offices in other Departments."

The first issue which arises is whether at the relevant time there was an equal employment opportunity program in the Department. Annexure "FCM 1" to the affidavit of Mr. Murray sworn 12 September 1988 was a bulky document which he described as the Equal Employment Opportunity Program of the Department, dated 1 October 1985. One of the stated objectives therein was to ensure that selection and promotion processes complied with "EEO principles" and that all promotion and selection committees observed "EEO requirements". A document was issued by the Department headed "EEO Guidelines for Selection Advisory Committees". The guidelines offered assistance, expressed didactically, as to the conduct of interviews and assessment of written applications made by, inter alia, women. It was designed to assist the implementation of the "equal opportunity principles" in the Program dated 1 October 1985.
The learned primary Judge held (84 ALR at 419-420) that these documents were insufficient to constitute a "program" because they would not be "legally enforceable". Accordingly, the first respondent failed at first instance upon this branch of her case.
However, his Honour does not appear to have had drawn to his attention the definition of "program" in sub-s. 22B (1) of the Public Service Act. This definition is inclusive and refers specifically, inter alia, to "the particular objectives to be achieved by the program" (para. (a)) and to "the policies to be adopted and the procedures to be followed, to achieve those objectives" (para. (b)). In our view, the document which is annexure "FCM 1" satisfied those elements in the definition of "program". It may be that those objectives, policies and procedures will be subject to further refinement and specification, but in their state as at October 1987, they fell within the statutory definition.
Further, with respect to his Honour, legal enforceability is the result of there being a program, not a constituent element in the definition of program. Indeed, and so far as sub-s. 22B (5) of the Public Service Act is concerned, "legal enforceability" has a dual character. A duty is imposed on the Secretary to take any action necessary to give effect to the program, and a duty is imposed on any person who exercises powers in relation to employment matters to have regard to the program in exercising those powers. One may "have regard to" objectives which are expressed in exhortative rather than specific terms.
In our view, there was at the relevant time a program within the meaning of sub-s. 22B (1) of the Public Service Act. This leads us to consider whether the appellant failed to comply with sub-s. 22B (5) in any way which founds a good claim by the first respondent for judicial review.
In his affidavit sworn 12 September 1988, Mr. Murray stated that the Department tried to ensure that due regard was had to the Equal Employment Opportunity Program. He went on to give various examples. One of the documents annexed to his affidavit was one to which we have already referred, that entitled "EEO Guidelines for Selection Advisory Committees". In his oral evidence, Mr. Murray said that that document was not used by posting advisory committees. Mr. Murray took the view that its use was required only in relation to decisions to promote rather than to transfer an officer. But the document itself envisages use in decisions for both promotions and transfers. Further, the definition of "employment matters" in sub-s. 22B (1) of the Public Service Act extends to both promotions and transfers. However, it is also apparent from Mr. Murray's oral evidence that in reaching the decision in question, he was well aware of what he described as "the EEO factor" and of the principles involved in the program represented by annexure "FCM 1".
In these circumstances, is the first respondent correct in her submission that there was relevantly a failure to have regard to the program and thus a failure in compliance with sub-s. 22B (5) of the Public Service Act?
There is a number of decisions construing legislation which obliged a decision maker to have regard to one or more particular matters in reaching a decision; see e.g. The Queen v Toohey; Ex parte Meneling Station Pty. Ltd. (1982) 158 CLR 327 at 333, 338 and cf. Rathborne v Abel (1964) 38 ALJR 293 at 294-295, 301, 303. The requirement in sub-s. 22B (5) of the Public Service Act to have regard to an equal employment opportunity program has to be read in the light of the somewhat generalised, exhortative and loosely expressed terms in which the legislation indicates what is included in the term "program". Further, as the present case illustrates, the program may be embodied or represented or evidenced by voluminous documentation. In our view, the "Guidelines" had been devised to assist decision makers in furtherance of the objectives in the program. A decision maker might have regard to the particular objectives and policies included in the program even though he misconceived the use to which the "Guidelines" document was designed to be put.
No doubt the substantial elements in a program, loosely expressed though they may be, must be taken into account and weight be given to the program as a whole as an element in the decision reached by the decision maker upon the particular case in question. Whether what is done in a particular case fairly answers this description will, of course, depend upon the facts. In the present case, in our view, even though Mr. Murray was mistaken as to the applicability of the particular document "EEO Guidelines for Selection Advisory Committees", that did not mean, looking at the relevant circumstances as a whole, that he did not have regard to the equal employment opportunity program. Accordingly, the point put in the first respondent's Notice of Contention fails.
Conclusions
In the result, of the four elements of the concept of sex discrimination presented by sub-s. 5 (2) of the Act, the first, second and fourth, but not the third, apply to the circumstances complained of by the first respondent. Accordingly, on this branch of the case, the appeal succeeds. Nor is the decision below saved by the point sought to be made good by the first respondent in her Notice of Contention.
By his Amended Notice of Appeal filed in Court on 25 May 1989, the appellant no longer seeks an order that the first respondent pay his costs either of the appeal or of the proceedings at first instance. Accordingly, in this Court, there should no order as to the costs of the appeal. The appeal should be allowed and in lieu of Orders 1, 2, 3 and 5 of the Orders made 18 October 1988, it should be ordered that the proceedings be dismissed. Order 4 of those Orders, which is to stand, provides for the costs below of the present first respondent to be paid by the present appellant.

JUDGE2

I have read the reasons of Bowen C.J. and Gummow J., which relieve me of the necessity of setting out the statutory provisions and facts relevant to the appeal in full.

As to the alleged breach of s.14(2) of the Sex Discrimination Act 1984, the principal issue is whether the definition of discrimination in s.5(2) of that Act applies. That, in turn, depends in the first place upon whether the discriminator "requires the aggrieved person to comply with a requirement or condition ... with which a substantially higher proportion of persons of the opposite sex to the aggrieved person comply or are able to comply". The section requires one to consider proportion, not number. In addition, there are two separate conditions, (b) and (c), further mentioned below.
The second edition of the Oxford English Dictionary gives as its second meaning of the word "proportion":

"The relation existing between things or magnitudes as to size, quantity, number or the like; comparative relation, ratio".

The first example, from the 14th century, gives the proportion of the circumference of a circle to its diameter as 22:7. This is plainly the sense in which the word "proportion" is used in s.5(2) - a relationship between numbers; such a relationship is often conveniently expressed as a fraction, e.g. one-half, or a percentage, e.g. 50%.
What is contemplated is that there will be a comparison between two proportions, to decide if one is substantially higher than the other. The words "substantially higher" obviously leave room for argument, but will not always do so. For example, if one-tenth of the relevant persons of one sex complied and one-thousandth of the relevant persons of the opposite sex did so, no one would doubt that the former (one-tenth) was a substantially higher proportion.
There is left a legal problem, namely to define the group or groups whose numbers constitute (to put it mathematically) the denominators, in the two fractions which are to be compared.
Bowen C.J. and Gummow J., following remarks of Street C.J. in Australian Iron & Steel Pty Ltd v. Najdovska (1988) 12 NSWLR 587 at 596 ("A.I.S.") have acted on the view that the denominator is the same for the men as for the women - i.e. the denominator is the total of the relevant men and women.
The result is that in such a case as A.I.S., if one-tenth of the women and one-tenth of the men in a factory are being put off, a court may still be obliged to hold that s.5(2)(a) is satisfied, if there are in the factory many more men than women, or vice versa. This construction makes no practical sense, since the intention is plainly to determine if the requirement or condition is one which, although it mentions neither men nor women, is "loaded" against one sex or the other. However, my rejection of the "joint base pool" construction is not founded mainly on that consideration, but on the language used: "proportion of persons of the opposite sex". This cannot ordinarily mean, as a matter of English, that the denominator consists of people of both sexes; the words "opposite sex" are singular, not plural. What the section requires is that you compare the proportion of persons of one sex who comply with the proportion of persons of the opposite sex who comply. The words after "of" in the expression "proportion of" indicate the denominator, in expressions such as: "The proportion of young people who marry is becoming less".

What the construction I have eschewed does is to compare, not two proportions, but two numbers; it compares the number of men with the number of women in a joint base pool and that is not what the section requires.
The next question is: which men and which women constitute the base pools? Putting that another way, who are to be numbered among the "persons of the opposite sex" who form one denominator and who are to be numbered among the persons, of the same sex as the aggrieved person, who form the other denominator?
The section simply does not say.
But the legislature's failure to give any express indication of the identity of the denizens of the base pools must be remedied by the courts, if the statute is to have any comprehensible effect at all. The only other course is to hold that the choice "of an appropriate section of the population is ... an issue of fact": Kidd v. D.R.G. (U.K.) Ltd (1985) IRLR 190 at 195 (para.20). It is not wholly such an issue, for there is room for rational disagreement as to what implication - for implication there must be - is to be made so that the section not fail for uncertainty.
The notion that all men and all women in the world constitute the two base pools is easily rejected, on the ground of absurdity. It is not credible that the legislature intended the courts to have regard, in deciding the legality of redundancy policies in a Sydney factory, to the question of what proportion of men and what proportion of women in, say, China would qualify for redundancy if they worked in the factory.
The majority in the A.I.S. case correctly (in my respectful opinion) accepted that there must be two separate denominators - one for each of the two proportions compared. But their Honours would include the whole of the men and the whole of the women within the jurisdiction. Priestley J.A. said:

"It seems to me that the groups which should be taken, since the Act is applicable in New South Wales, are all men and all women in New South Wales" (p 610E).

His Honour did not explain the reason for the limitation to New South Wales, but it must be based upon the presumption against extra-territorial operation. Mason J. (as his Honour then was) said of hire-purchase legislation in Goodwin v. Jorgensen (1973) 128 CLR 374 at 383:

"Section 41 would exceed the legislative power of the State if it were construed as creating an offence anywhere in the world. To preserve its validity it should be read subject to some qualification. The qualification which obviously presents itself is that the section applies to act and omissions taking place in New South Wales."

The most obvious qualification, as to anti-discrimination statutes, is to limit their operation to acts of discrimination, such as advertising, taking place within the jurisdiction. But it is by no means clear that such a statute should be, prima facie, construed as relevant only to people who happen to be within the jurisdiction when the act of discrimination takes place. Whether or not it should be so construed, I do not see any ground for the implication that such a statute, in dealing with indirect discrimination, takes as its base pools the maximum groups of people to whom the statute can apply. Not that Priestley J.A. does precisely this; in writing of "all men and all women in New South Wales", his Honour presumably intends to exclude infants. I suggest that the view of Street C.J. in the A.I.S. case at p 596 is correct: "the base group is to be constituted by the persons to whom the requirement or condition was directed and who are divided or segregated by it", with the reservation that I would, as explained above, take two separate base groups or pools.
Thus, in looking at a condition (e.g. age) by which those who are to stay are to be segregated from those to be dismissed, one does not have to consider the condition's application to men and women not employed in the factory in which the condition has been applied. If the condition catches one woman and ten men, then you consider only the women and men in the factory, to determine whether it applies to a substantially higher proportion of women than men or vice versa; should there be, in total, 10 women and 100 men in the factory, then there is no indirect discrimination. On the test used by Street C.J., one would compare the proportion of women affected by the condition to the total number of people in the factory (1:110) with the proportion of men so affected to the total number of people in the factory (10:110), which as I have said simply amounts to comparing the numbers affected. The test of Priestley J.A. would require one to embark on the stranger task of considering the effect of the redundancy condition on people who do not work in the factory and to whom the condition is, therefore, irrelevant.
Unfortunately, in the present case, reaching a view on these basic questions of construction does not solve the entire problem, for there is much room for difference of view about the size of the group of men who are divided or segregated by the requirement as to grade and the size of the group of women who are so divided or segregated; compare the difficulty encountered by Schiemann J. in R. v. Secretary of State for Education; Ex parte Schaffter (1987) IRLR 53 (para.25). Each base pool must be set at such a size as to exclude those who are only theoretically affected by the requirement or condition; otherwise, requirements or conditions having the effect of discriminating may, by the use of unrealistically composed base pools, escape sanction. As to selection of personnel for appointment or transfer, then, one must confine the base pools to those who, but for the impugned requirement or condition, might rationally be regarded as entitled to consideration.
Perhaps because of the difficulty of construction of s.5(2)(a), the evidence as to the number of persons in the base pools was deficient. An attempt was made to remedy this at and after the hearing in this Court. There was, in the end, no consensus as to the relevant figures. According to statistics relied upon by the appellant, if one treated the base pools as consisting of those journalists in the Department at the relevant time who were "substantive occupants" positions in "Promotion Australia" (the relevant segment of the Department) 13.6% of the men and 8.3% of the women could comply. If one considered journalists within the Department as a whole, 28.8% of the men and 21.4% of the women could comply. However, the figures just mentioned are not conceded by the first respondent to be accurate, and it does not appear, therefore, that it would be proper to rely upon them.
The question whether the difference between the pairs of figures just mentioned are substantial, in the relevant context, is essentially one of fact, and I do not think this Court should try to decide it. The difficulty is that, although there was a great deal of material before the trial judge, little attention appears to have been paid by the parties to the critical initial question of selection of the base pools. His Honour was apparently invited to, and did, proceed on the view that the question was whether a substantially higher number of men than of women were able to comply; that is, plainly, not what the statute intends. The "base pool" for his Honour was a group consisting of both men and women, and the group was made up of all those able to comply; on my construction of the Act, the judge was in error on both points.
It is my opinion that this Court should not send the matter back for a full, or limited, rehearing. It is unfortunate that, due to the extraordinarily vague drafting of the indirect discrimination provision an error has (in my respectful opinion) been made below in applying it. Particularly in view of the time which has elapsed, it would seem quite impractical to attempt to remedy that now; my view is that the first respondent should be held to have failed on the indirect discrimination point, on the ground of lack of proof of the foundation of that part of the case, namely that the proportion of relevant men who could comply with the condition substantially exceeded the proportion of relevant women who could do so.
This conclusion makes it unnecessary, strictly speaking, to consider the other questions as to indirect discrimination, but it is desirable to deal with one point which was argued.
As Bowen C.J. and Gummow J. point out in their reasons, some of the English cases support the view that a "requirement or condition" must be an "absolute bar". Having considered the English cases, and in particular Clarke v. Eley (IMI) Kynoch Ltd (1982) IRLR 482 at para.12; Watches of Switzerland Ltd v. Savell (1983) IRLR 141 at para.23; and Perera v. The Civil Service Commission (1983) IRLR 166 at para.25, I agree with the conclusion of Bowen C.J. and Gummow J. that:

"A requirement or condition, therefore, means a stipulation which must be satisfied if there is to be a practical (and not merely a theoretical) chance of selection."

It follows, in my opinion, that the learned primary judge was correct in holding that there was a requirement or condition in the present case.
Equal Employment Opportunity Program
I agree with Bowen C.J. and Gummow J. that such a program exists.
The definition of the expression set out above is included in the reasons of Bowen C.J. and Gummow J. It will be noticed that the definition is framed in terms of the purpose which the program is designed to achieve; the purpose stipulated is, to put it shortly, to ensure the taking of appropriate action and measures to eliminate certain unjustified discrimination, and to equalise the career opportunities of certain people, namely women and "persons in designated groups". Then there is a separate definition of the single word "program", which is also set out in the reasons of Bowen C.J. and Gummow J. I note, in particular, that the word includes the program's "particular objectives" and:

"(b) the policies to be adopted, and the procedures to be followed, to achieve those objectives".

Section 22B(5) of the Public Service Act 1922 reads as follows:

"The Secretary of a Department shall take any action necessary to give effect to the equal employment opportunity program for the Department and any person who exercises powers in relation to employment matters in the Department shall have regard to the program in exercising those powers".

Insofar as the program sets out policies to be adopted, it is easy to understand what it means to require people exercising relevant powers to have "regard to" the program, as contemplated by ss.22B(4) and (5). But insofar as the program sets out "the procedures to be followed", it is unclear whether the intention was that those procedures would be binding - i.e. would have the force of law. Section 22B(5) appears to intend a contrast between the Secretary's obligation, namely to take any action necessary to give effect to the program, and the obligation of the other persons mentioned in the sub-section, which is only to "have regard to" the program. The latter expression is not ordinarily appropriate, if the intention is to make absolutely binding legal rules; it is not ordinarily equivalent to "comply with".
Another preliminary observation which may be made is that it is a question whether the obligation to have regard to the program obliges the persons who are subject to it to have regard also to policies or procedures which are not set out in the program itself, but are set out in documents composed by persons purporting to act under the authority of the program. More shortly, it is unclear whether the program maker's authority may be delegated. The matter was not argued before us, but in my opinion, at least as far as the part of the definition of "program" lettered (b) is concerned (quoted above), the intention is that the policies and procedures shall be set out in the program itself. It is my view that the Secretary cannot comply with his obligation (created by s.22B(3)) to prepare a statement setting out the "program" as defined by merely nominating other persons to specify the policies to be adopted and the procedures to be followed; the definition of "program" appears to contemplate that those matters will be set out in the program itself. I appreciate that, in the course of carrying out objectives or policies stated in a program, directions (oral and in writing) would be given at various levels of a Department; but they do not, in my opinion, become part of the program. Such directions do not derive their legal force from s.22B(5), quoted above.
What I have said on the point just mentioned may be illustrated by the "Guidelines", allegedly part of the program, referred to in the reasons of the learned primary judge (84 ALR 408 at p 420) and further mentioned below. His Honour left aside "the difficult question whether the guidelines constitute a part of the 'program', whose observance is required by s.22B(5)". I find it necessary to express a view on that point. It is that the Guidelines, not being part of the program prepared under s.22B(3), are not matters to which regard must be had under s.22B(5). I therefore agree with his Honour that the forms set out in the Guidelines are not part of the program.
The Guidelines, whose full title is "EEO Guidelines for Selection Advisory Committees" were apparently brought into existence under clause C2.5 of the program, which reads in part: "Devise criteria directing promotion/selection committees to take into account the quality of applicants ..." It follows from what I have said above that in my view such criteria, when devised, would form no part of the program. This point is of some importance, not only because it was dealt with by the primary judge, but because s.33 of the Public Service Act 1922 sets out, apparently comprehensively, the criteria to be used in making appointments; for example, patronage and favouritism are proscribed. Sub-section 4(b) preserves the effect of:

"a program to encourage the appointment of women or persons in a designated group to the Service, being a program that is declared by the regulations to be an approved program for the purposes of this paragraph".

We were referred to no such regulations; but in any event I do not accept that Parliament intended that the statutory criteria set out in s.33 should be able to be supplemented, or perhaps varied, by criteria not directly contained in any program, but prepared under the program's authority. Section 33 is a provision of central importance in the Act.

The primary judge found, in effect, that the relevant officer (Mr Murray) had not used the Guidelines and held, rightly in my opinion, that he was not obliged to use the forms in them. It is desirable to set out the reasons in my view which justified that conclusion:
1. Mr Murray was obliged to have regard to the program by

s.22B(5); that provision did not oblige him to have regard to requirements promulgated in reliance on authority purporting to have been given by the program.

The Guidelines themselves (annexure FCM9) do not require the

use of any forms.

If the forms were included in the program (as they are not)

it would make little sense to ask whether Mr Murray had "had regard to them". The forms are intended for use; one either uses them or one does not.

As to the last point, I feel obliged to say that one can understand, and to an extent sympathise with, reluctance to use the forms. They contemplate that all the questions asked will be written down - an onerous requirement, at least in a long interview - and that no applicant will be asked a question not asked of others.
Some other points were taken by the first respondent about the program, but none of them had any substance, in my view.
In my opinion, the appeal should be allowed, and the order of the learned primary judge set aside. I would not allow any costs, here or below, in view of the fact that the obscurity of the relevant provisions is not the appellant's fault and because of the importance of the issues raised; in any event, no orders for costs are sought by the appellant.

[1989] FCA 546

28 AUGUST 1989