FEDERAL COURT OF AUSTRALIA

Fisher v Commonwealth of Australia [2023] FCAFC 106

ORDERS

THE COURT ORDERS THAT:

1.    The questions of law in the amended special case set out in Annexure B of the orders of the docket judge dated 9 December 2022 be answered as follows:

(1)    Do the Applicant and each of the represented persons have the same interest in the proceeding, save for the relief set out in paragraph 2 of the amended originating application?

Answer: Yes.

(2)    Do:

a.    the Applicant; and, or alternatively

b.    the represented persons,

enjoy the right to apply for and receive the age pension “to a more limited extent” (within the meaning of that expression in s 10 of the Racial Discrimination Act 1975 (Cth)) than non-Indigenous men born on or between 1 January 1957 and 31 December 1957, by reason of:

c.    ss 23(5A) and 43 of the Social Security Act 1991 (Cth); or

d.    alternatively, s 3 and Schedule 11, item 1, of the Social Security and Other Legislation Amendment (Pension Reform and Other 2009 Budget Measures) Act 2009 (Cth) (as it applied to item 5 of the table in s 23(5A) of the Social Security Act)?

Answer: No.

(3)    If the answer to Question 2 is yes, does s 10 of the Racial Discrimination Act operate such that the “pension age” for:

a.    the Applicant; and, or alternatively

b.    the represented persons,

for the purposes of item 5 of the Table appearing in s 23(5A) of the Social Security Act, and for the purposes of s 43 of the Social Security Act, is:

c.    64 years of age?

d.    Any other age less than 67 years of age?

Answer: Does not arise.

2.    Within 35 days, each party file and serve a short written submission on any further orders that they contend should be made by the Full Court.

3.    Subject to order 2, the proceeding otherwise be referred back to a docket judge for case management.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

THE COURT:

1     INTRODUCTION

1    The applicant is an Aboriginal man who was born on 23 June 1957 and is now 66 years old.

2    Pursuant to s 43(1) of the Social Security Act 1991 (Cth) (SSA), a person qualifies to receive the age pension when they reach the “pension age”, if one of four defined circumstances is present. The applicant meets the criterion in s 43(1)(a), having been resident in Australia for more than 10 years.

3    The “pension age” is defined by s 23(1) of the SSA. For a man born after 1 January 1957 it is 67 years. Accordingly, if the SSA is applied according to its terms, the applicant will qualify to receive the age pension on 23 June 2024. Whether a pension actually becomes “payable” to him will depend, under s 44, on whether his “pension rate” is greater than nil. Under s 55, the pension rate is worked out using either Pension Rate Calculator A (which appears at the end of s 1064) or Pension Rate Calculator B (which appears at the end of s 1065).

4    The applicant lodged a claim for the age pension on 9 September 2021. It was refused and no pension has been paid to him. He argues that, despite the terms of the provisions relating to the “pension age” in the SSA, he is entitled to be treated as meeting the qualification criteria for the age pension. This result is said to flow from the shorter life expectancy that Aboriginal men have, compared to other men in Australia, and the application of s 10(1) of the Racial Discrimination Act 1975 (Cth) (RDA). Put shortly, he says that s 10(1) requires that Aboriginal people, if they otherwise meet the relevant criteria in the SSA, receive the age pension for the same duration as other people. In broad terms, that is said to mean that an Aboriginal man born in 1957 qualified for the age pension when he reached 64 years of age. The applicant seeks declarations and an injunction to vindicate that position.

5    The proceeding is currently constituted as a representative proceeding under r 9.21 of the Federal Court Rules 2011 (Cth) (Rules) and some of the issues between the parties relate to whether it should continue to be constituted in that way. Again in broad terms, the represented persons are Indigenous (that is, Aboriginal and Torres Strait Islander) men who turned 65 in 2022 and who, if the applicable pension age is 64 rather than 67, meet the qualification and payability criteria and are thus entitled to receive a pension. (Aboriginal women are not included in the group of represented persons; however, if the applicant’s argument is upheld, it will be able to be relied on to some degree by them as well. The pension age for women in the relevant age group is the same (67 years) but the relevant life expectancy statistics are not the same.)

6    On 9 December 2022, pursuant to r 38.01 of the Rules, Mortimer J (as her Honour then was) referred an amended special case for the consideration of the Full Court (the special case). The special case states the following questions of law:

Question 1

Do the Applicant and each of the represented persons have the same interest in the proceeding, save for the relief set out in paragraph 2 of the amended originating application?

Question 2

Do:

(a)    the Applicant; and, or alternatively

(b)    the represented persons,

enjoy the right to apply for and receive the age pension “to a more limited extent” (within the meaning of that expression in s 10 of the Racial Discrimination Act 1975 (Cth)) than non-Indigenous men born on or between 1 January 1957 and 31 December 1957, by reason of:

(c)    ss 23(5A) and 43 of the Social Security Act 1991 (Cth); or

(d)    alternatively, s 3 and Schedule 11, item 1, of the Social Security and Other Legislation Amendment (Pension Reform and Other 2009 Budget Measures) Act 2009 (Cth) (as it applied to item 5 of the table in s 23(5A) of the Social Security Act)?

Question 3

If the answer to Question 2 is yes, does s 10 of the Racial Discrimination Act operate such that the “pension age” for:

(a)    the Applicant; and, or alternatively

(b)    the represented persons,

for the purposes of item 5 of the Table appearing in s 23(5A) of the Social Security Act, and for the purposes of s 43 of the Social Security Act, is:

(c)    64 years of age?

(d)    any other age less than 67 years of age?

7    The special case sets out over the course of 98 paragraphs the facts which the parties have agreed for the purpose of deciding these questions. A significant body of documentary material is annexed. The key points are these:

(a)    According to life tables compiled by the Australian Bureau of Statistics based on data for 2015 to 2017, an Indigenous man aged 65 had a remaining life expectancy of 15.8 years. A non-Indigenous man of the same age had a life expectancy of 19.0 years (special case at [24]).

(b)    From 2006 to 2018, mortality rates for Indigenous people improved at a similar rate to mortality rates for non-Indigenous people, so that the gap in life expectancy has not narrowed (at [26]).

(c)    “The reason for the shorter life expectancies of Aboriginal and Torres Strait Islander Australians is … that they experience greater adverse health outcomes compared to non-Indigenous Australians. Those outcomes are shaped by a range of interconnected structural, social and cultural determinants of health, including the historical and ongoing consequences of colonisation. As such, the gap in life expectancy between Aboriginal and Torres Strait Islander Australians and their non-Indigenous counterparts is a function of race.” (At [29]).

8    The gap in life expectancy arising from the figures in [7(a)] above is the one that is of direct relevance to the resolution of this proceeding. It is worth noting, however, that this reflects the statistical expectations for men who have survived all of the risks associated with life and reached the age of 65. As reflected in the same tables, life expectancy at birth was 71.6 years for an Indigenous male and 80.2 years for a non-Indigenous male. That reflects greater rates of mortality for Indigenous males at all stages of life.

9    The expression “a function of race” is somewhat imprecise and needs elaboration. As the succeeding paragraphs of the special case explain, Indigenous Australians suffer to a greater degree than others from a range of physical and mental health problems which contribute to mortality rates and thus life expectancy. The facts agreed by the parties do not suggest that these disparities arise from something inherent in Indigenous people or their cultures that make them inherently likely to live shorter lives than other people. Rather, to the extent that underlying causes are identified, those causes are connected to the ongoing effects of colonisation, dispossession, destruction of cultural bonds, poor access to services and racist policies. Thus, the gap in life expectancy is “a function of race” in the sense that it is (so far as the facts disclose) the product of disadvantages suffered by Indigenous Australians which, in turn, flow from their treatment by governments and by more powerful or fortunate Australians.

10    These facts, which the Australian government accepts for the purposes of the proceeding to be true, are a matter of grave concern for a society that values equality of opportunity. For that reason, as well as to give proper context to the short summary above, the statement of agreed facts contained in the special case is extracted (without its own annexure marked SC-01) and included as an annexure to these reasons.

2     SECTION 10 OF THE RDA

11    The key provisions of Pt II of the RDA are as follows:

9    Racial discrimination to be unlawful

(1)    It is unlawful for a person to do any act involving a distinction, exclusion, restriction or preference based on race, colour, descent or national or ethnic origin which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of any human right or fundamental freedom in the political, economic, social, cultural or any other field of public life.

(1A)    Where:

(a)    a person requires another person to comply with a term, condition or requirement which is not reasonable having regard to the circumstances of the case; and

(b)     the other person does not or cannot comply with the term, condition or requirement; and

(c)     the requirement to comply has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, by persons of the same race, colour, descent or national or ethnic origin as the other person, of any human right or fundamental freedom in the political, economic, social, cultural or any other field of public life;

the act of requiring such compliance is to be treated, for the purposes of this Part, as an act involving a distinction based on, or an act done by reason of, the other person’s race, colour, descent or national or ethnic origin.

(2)    A reference in this section to a human right or fundamental freedom in the political, economic, social, cultural or any other field of public life includes any right of a kind referred to in Article 5 of the Convention.

…

10        Rights to equality before the law

(1)    If, by reason of, or of a provision of, a law of the Commonwealth or of a State or Territory, persons of a particular race, colour or national or ethnic origin do not enjoy a right that is enjoyed by persons of another race, colour or national or ethnic origin, or enjoy a right to a more limited extent than persons of another race, colour or national or ethnic origin, then, notwithstanding anything in that law, persons of the first‑mentioned race, colour or national or ethnic origin shall, by force of this section, enjoy that right to the same extent as persons of that other race, colour or national or ethnic origin.

(2)    A reference in subsection (1) to a right includes a reference to a right of a kind referred to in Article 5 of the Convention.

…

12    Sections 9 and 10 are qualified by s 8, which is (relevantly) as follows:

8      Exceptions

(1)    This Part does not apply to, or in relation to the application of, special measures to which paragraph 4 of Article 1 of the Convention applies except measures in relation to which subsection 10(1) applies by virtue of subsection 10(3).

…

13    The Convention, referred to in these provisions, is the International Convention on the Elimination of All Forms of Racial Discrimination, opened for signature 21 December 1965, 660 UNTS 195 (entered into force 4 January 1969) (the Convention), the text of which is contained in the Schedule to the RDA. Its key provisions are as follows.

(a)    Paragraph 1 of art 1 defines the concept of “racial discrimination” to mean:

In this Convention, the term “racial discrimination” shall mean any distinction, exclusion, restriction or preference based on race, colour, descent, or national or ethnic origin which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of human rights and fundamental freedoms in the political, economic, social, cultural or any other field of public life.

(b)    Paragraph 4 of art 1 (which forms the basis for s 8 of the RDA) provides:

Special measures taken for the sole purpose of securing adequate advancement of certain racial or ethnic groups or individuals requiring such protection as may be necessary in order to ensure such groups or individuals equal enjoyment or exercise of human rights and fundamental freedoms shall not be deemed racial discrimination, provided, however, that such measures do not, as a consequence, lead to the maintenance of separate rights for different racial groups and that they shall not be continued after the objectives for which they were taken have been achieved.

(c)    By art 2, the States Parties condemn racial discrimination and undertake to pursue a policy of eliminating it. Relevantly for present purposes, para 1(c) of art 2 provides:

Each State Party shall take effective measures to review governmental, national and local policies, and to amend, rescind or nullify any laws and regulations which have the effect of creating or perpetuating racial discrimination wherever it exists;

…

(d)    In art 5, the States Parties undertake to do two things: to prohibit and eliminate “racial discrimination in all its forms” (the basis for s 9 of the RDA); and (relevantly to s 10):

… to guarantee the right of everyone, without distinction as to race, colour, or national or ethnic origin, to equality before the law, notably in the enjoyment of the following rights:

(a)    The right to equal treatment before the tribunals and all other organs administering justice;

(b)    The right to security of person and protection by the State against violence or bodily harm, whether inflicted by government officials or by any individual, group or institution;

(c)    Political rights, in particular the rights to participate in elections – to vote and to stand for election – on the basis of universal and equal suffrage, to take part in the Government as well as in the conduct of public affairs at any level and to have equal access to public service;

(d)    Other civil rights, in particular:

(i)    The right to freedom of movement and residence within the border of the State;

(ii)    The right to leave any country, including one’s own, and to return to one’s country;

(iii)    The right to nationality;

(iv)    The right to marriage and choice of spouse;

(v)    The right to own property alone as well as in association with others;

(vi)    The right to inherit;

(vii)    The right to freedom of thought, conscience and religion;

(viii)    The right to freedom of opinion and expression;

(ix)    The right to freedom of peaceful assembly and association;

(e)     Economic, social and cultural rights, in particular:

(i)    The rights to work, to free choice of employment, to just and favourable conditions of work, to protection against unemployment, to equal pay for equal work, to just and favourable remuneration;

(ii)    The right to form and join trade unions;

(iii)    The right to housing;

(iv)    The right to public health, medical care, social security and social services;

(v)    The right to education and training;

(vi)    The right to equal participation in cultural activities;

(f)     The right of access to any place or service intended for use by the general public such as transport, hotels, restaurants, cafes, theatres and parks.

2.1     Section 10: Introductory observations

14    Section 10 of the RDA differs from s 9 in two important ways. First, while s 9 prohibits conduct, s 10 is directed at the operation of laws. Section 10 therefore cannot be contravened in the way that s 9 can. Secondly, while s 9 is expressly based on the concept of discrimination (including what is often referred to as indirect discrimination, which is the subject of s 9(1A)), s 10 is not. The latter section is expressed to apply where there is an unequal enjoyment, as between members of different races, of a relevant “right”. That does not mean that concepts of discrimination are irrelevant, however. Section 10 sits alongside s 9 in a statutory regime squarely directed at that phenomenon. Both provisions use as the touchstone for their operation the impairment of the enjoyment of basic or fundamental human rights, in circumstances where that impairment is connected, directly or indirectly, to race: see Mabo v Queensland [1988] HCA 69; 166 CLR 186 at 216–217 (Brennan, Toohey and Gaudron JJ) (Mabo No 1).

15    Section 10, applying as it does to a law of the Commonwealth or a State or Territory, can have different effects in different circumstances.

(a)    Where a law of a State limits or partially abrogates a pre-existing domestic law right in a way that leads to the unequal enjoyment of a human right as between members of different races, it is inconsistent with s 10 and is therefore rendered inoperative (in whole or in part) by s 109 of the Constitution (eg Gerhardy v Brown [1985] HCA 11; 159 CLR 70 at 98–99 (Mason J) (Gerhardy); Western Australia v Ward [2002] HCA 28; 213 CLR 1 at [107]–[108] (Gleeson CJ, Gaudron, Gummow and Hayne JJ) (Ward)). The same consequence occurs in respect of a law enacted by a Territory legislature, by force of provisions of Territory self-government legislation that give Commonwealth Acts primacy (see, eg, Bara v Blackwell [2022] NTCCA 17 at [20] (Bara)).

(b)    If a law of the Commonwealth enacted after 1975 is said to abrogate or restrict a right inconsistently with s 10, issues will arise as to whether it thereby effects an implied partial repeal (cf, eg, Ward at [99]).

(c)    On the other hand, where a law (of the Commonwealth or a State or Territory) confers or expands a right which gives effect to a relevant human right, but omits to do so universally, s 10(1) has the effect of extending or augmenting that conferral to the extent necessary to eliminate the inequality of its enjoyment between people of different races. In Gerhardy at 98 (in a passage adopted by the majority in Ward at [106]) Mason J said:

If racial discrimination arises under or by virtue of State law because the relevant State law merely omits to make enjoyment of the right universal, i.e. by failing to confer it on persons of a particular race, then s. 10 operates to confer the right on persons of that particular race.

16    It is the third area of s 10’s operation that is engaged here, if the applicant’s submissions are accepted.

17    In very broad terms, the applicant submits that s 10(1) is engaged whenever a law confers a right in a way that results in members of different races enjoying the relevant human right to different extents. That includes the case where a law applies a neutral criterion but, because members of a particular race find themselves in different circumstances to other members of the community, they do not meet or are less likely to meet that criterion. The Commonwealth submits that such a far-reaching operation cannot have been intended and that s 10 applies in a more limited way: when the law is expressly framed so as to apply distinctly or differentially to members of a particular race; or when the law is facially neutral but, when seen in context, is found to target members of a particular race.

18    Before turning to the issues of construction that arise, five general points should be made about s 10(1).

19    The first general point is that the purpose of the RDA — and in particular the substantive provisions in Pt II — is to implement the Convention. Its provisions are to be construed accordingly. To the extent that the language of the RDA permits, it is to be construed so as to be consistent with the Convention: Acts Interpretation Act 1901 (Cth) ss 15AA and 15AB; Minister for Immigration and Multicultural and Indigenous Affairs v QAAH of 2004 [2006] HCA 53; 231 CLR 1 at [34]. It must be borne in mind, however, that it is the RDA that forms part of the law of Australia and not the Convention itself (see, eg, Maloney v the Queen [2013] HCA 28; 252 CLR 168 at [174] (Kiefel J) (Maloney); Dietrich v The Queen [1992] HCA 57; 177 CLR 292 at 305 (Mason CJ and McHugh J), 359–360 (Toohey J)).

20    No question concerning the validity of the RDA arises in this case. However, it is useful to note the basis upon which the Act has been regarded as within the Commonwealth’s legislative power, as it has some relevance for the task of construction. Sections 9 and 12 of the RDA were upheld as valid enactments under s 51(xxix) of the Constitution (the external affairs power) in Koowarta v Bjelke-Petersen [1982] HCA 27; 153 CLR 168 (Koowarta). As described more recently in Maloney at [61] (Hayne J), the holding in Koowarta is that the RDA is a valid enactment because it implements Australia’s obligations under the Convention. This observation emphasises the need for provisions of the RDA to be construed consistently with the Convention.

21    Although the Preamble to the RDA suggests an attempt to rely on “all relevant powers of the Parliament”, including s 51(xxvi) and s 51(xxvii) of the Constitution (“the people of any race for whom it is deemed necessary to make special laws” and “immigration and emigration”), the attempt to rely on the races power as an additional foundation seems to have been unsuccessful (Koowarta at 187 (Gibbs CJ), 211 (Stephen J), 244–245 (Wilson J), 261–262 (Brennan J)). It does not appear to have affected the understanding, reflected in the authorities, that the purpose of the RDA is to implement the Convention. Section 10 itself has been described as intending to comply with or implement art 2.1(c), art 5 or both: see Gerhardy at 95 (Mason J); Viskauskas v Niland [1982] HCA 15; 153 CLR 280 at 294 (Gibbs CJ, Mason, Murphy, Wilson and Brennan JJ); Mabo No 1 at 217 (Brennan, Toohey and Gaudron JJ); Maloney at [10] (French CJ), [161] (Kiefel J), [201] (Bell J), [303] (Gageler J).

22    The second general point is that s 10(1) is a remedial statute directed, at a fundamental level, to the protection of human dignity. It is therefore “not to be given a legalistic or narrow interpretation” (Mabo No 1 at 230 (Deane J); Western Australia v Commonwealth (Native Title Act Case) [1995] HCA 47; 183 CLR 373, 437 (Mason CJ, Brennan, Deane, Toohey, Gaudron and McHugh JJ) (Native Title Act Case)). Issues concerning the effect of laws and the equal or unequal enjoyment of relevant rights should therefore be approached as matters of substance rather than form (Maloney at [38] (French CJ), [65], [84] (Hayne J (Crennan J agreeing at [112])), [148] (Kiefel J), [204] (Bell J), [343(a)] (Gageler J)).

23    The third general point (which flows from the first point) is that the “rights” of which s 10(1) speaks are fundamental human rights. Section 10(2) directs attention to the rights listed in art 5 of the Convention. The list of rights in art 5 is introduced by the word “notably” and is therefore itself non-exhaustive. The field of operation of the Convention is identified by art 1(1), which refers to the enjoyment or exercise on an equal footing of “human rights and fundamental freedoms in the political, economic, social, cultural or any other field of public life”.

24    Despite the non-exhaustive terms of s 10(2), the “rights” to which s 10 relates have been understood to comprise human rights of the kind referred to in the Convention (eg Mabo No 1 at 216–217 (Brennan, Toohey and Gaudron JJ); Maloney at [9] (French CJ), [63] (Hayne J), [145] (Kiefel J), [300] (Gageler J)). The constitutional underpinning of the RDA and its adherence to the task of implementing the Convention indicate that the expression “right” should be taken to refer to the human rights with which the Convention is concerned, not to particular legal rights existing from time to time under domestic law.

25    The fourth general point is that “race” is a complex and in some ways problematic concept. Section 10, like the Convention, seeks to minimise such complexities by referring to “race, colour or national or ethnic origin”. While concepts of racial discrimination are well understood, identification of a particular community as a “race”, let alone who is or is not a member of it, can sometimes be fraught. However, in the context of s 10, the term is “not to be given a pedantic or narrow meaning” (Mabo No 1 at 230 (Deane J)). There is no dispute in the present case that Aboriginal Australians constitute a “race” for relevant purposes or that the applicant, and the represented persons, are members of that race. In what follows, therefore, it is possible to refer in a general way to people of a “race” without going into the complexity attending that terminology.

26    The fifth general point — albeit an obvious one — is that the ordinary canons of statutory construction apply. Effect must be given to the text enacted by Parliament, while all relevant aspects of context and accepted constructional rules must be considered in ascertaining its meaning. The section should be read as a whole, as a single enactment; the meaning of particular words or phrases can only be understood in their interaction with the rest of the section. And the section should be read together with other provisions of the RDA so as to operate, to the extent possible, in harmony with them.

27    Before going to the cases in more detail, it should be noted that the present case raises issues concerning the operation of s 10 that have received very little attention beyond the short statement by Mason J in Gerhardy set out above at [15] (and its endorsement in Ward). Arguments based on s 10 have succeeded in only a small number of cases, and all of these have involved State laws that abrogated or restricted pre-existing rights. The result was invalidity, subject to s 8. Courts have not yet had to deal in detail with questions as to how s 10 operates in the case of a statute that gives effect to a human right by creating or conferring rights under Australian law, but does so in a way that leads to unequal enjoyment of the relevant human right. These questions give rise to some complexity, which we discuss later in these reasons.

28    It is also appropriate to observe that, where the impugned law is a law of the Commonwealth made after the RDA, the issues cannot be avoided by concluding that the law is in conflict with s 10 and thus effects a partial repeal. The Commonwealth properly did not advance any submission of this kind. In truth there is no conflict, because (according to Gerhardy and Ward) s 10 confers rights in addition to, rather than in conflict with, the impugned law.

2.2     Section 10: “by reason of … a law”

29    Significant energy was expended in the Commonwealth’s submissions in attempting to construe the phrase “by reason of … a law” in s 10(1). We found this exercise unpersuasive, not only because the section must be read as a whole (cf, in relation to s 9, Wotton v Queensland (No 5) [2016] FCA 1457; 352 ALR 146 at [530] (Mortimer J)). Where the relevant human right is not capable of enjoyment except as a result of the conferral by an Australian statute of particular legal rights (as is the case here), it seems to us impossible as a matter of ordinary English to say that any unequal enjoyment of that human right does not occur “by reason of” that statute. The unequal enjoyment is necessarily a consequence of the choices that the Parliament made in designing the way in which rights would be conferred.

30    Cases such as Aurukun Shire Council v Chief Executive Officer, Office of Liquor Gaming and Racing [2010] QCA 37; [2012] 1 Qd R 1 (Aurukun), Munkara v Bencsevich [2018] NTCA 4 (Munkara) and Bara (which are discussed below) therefore have little if anything to say about this aspect of the present case. These cases involved laws that, to the extent they affected the enjoyment of human rights, were restrictive of pre-existing rights. In each case the impugned law was facially neutral and it was held that, to the extent that it affected Aboriginal persons to a greater extent than other members of the community, that was a result of social conditions or individual conduct rather than “by reason of” the law. Arguments were put in the present case as to whether these decisions could stand with the reasons of members of the High Court in Maloney. Regardless of the answer to that question, the reasoning in those cases does not support the conclusion that the unequal enjoyment of a right created by a statute does not occur “by reason of” that statute.

31    It also follows from what we have said above that we respectfully doubt the correctness of the reasoning of Goldberg and Hely JJ in Sahak v Minister for Immigration [2002] FCAFC 215; 123 FCR 514 (Sahak) at [49]. Sahak (which is also discussed below) concerned the limitation period imposed by a provision of the Migration Act 1958 (Cth) (Migration Act) for making an application for judicial review in this Court. The provision was part of a regime conferring, and regulating access to, review rights. To the extent that those rights were enjoyed to a lesser degree by members of particular races (which was said to be the case because English was not their first language), we doubt whether that could properly be said not to be a consequence of how the review regime was designed, including as it did a relatively short and non-extendable limitation period.

32    The Commonwealth submitted that s 10(1) is engaged only where the law in question either is expressed to apply differentially on the basis of race (including by reference to rights or interests that attach to members of a race, such as native title) or is found to have adopted a facially neutral criterion as a conscious proxy for race. In light of s 10(1)’s concern with substance rather than form, and the need to give its terms a generous and non-technical construction, we do not think that the phrase “by reason of … a law” can be narrowed in that way. Internationally and in Australia, it has long been recognised that discrimination can sometimes be found in the equal treatment of that which is unequal: eg Gerhardy at 129 (Brennan J). Thus, where a legislative choice is made to apply a neutral criterion that has an unequal practical effect, it is no answer to say that such an effect does not occur by reason of the law.

2.3     Section 10: Unequal enjoyment of rights

33    A more useful (but also more complicated) question is what, in context, is meant by the reference to persons of a particular race not “enjoying” a right enjoyed by others or “enjoying” it “to a more limited extent”. These expressions lie at the heart of s 10(1) and their meaning, to a large extent, defines its reach. Is there unequal “enjoyment” of a human right, as between members of a particular race and other persons, if Australian law gives force to that right by way of racially neutral criteria but it can be shown (including by statistical analysis) to be in fact less accessible, or of less utility, to members of that race than to others? Or is the concept (as the Commonwealth submitted) limited to cases where the human right is conferred or limited in a discriminatory way, whether expressly or by adoption of some apparently neutral criterion as a proxy for race?

34    While accepting that s 10 was concerned with matters of substance rather than form, the Commonwealth emphasised that decisions of the High Court have so far found s 10 to be engaged only where the law in question singled out people of a particular race, either by expressly applying to those persons (or rights attaching to those persons) or by being, despite a facially neutral criterion, targeted at those persons.

(a)    Gerhardy concerned a provision in South Australian legislation which prohibited any non-Pitjantjatjara person from entering a large tract of the State without permission. ‘Pitjantjatjara’ was defined to mean a person who is a member of particular groups and has traditional ownership rights in the land. The provision was held to be a “special measure” under s 8(1) of the RDA, but five Justices held that s 10(1) would otherwise have applied. The criterion of traditional ownership was regarded as distinguishing on the basis of race.

(b)    Mabo No 1 concerned a provision of Queensland law which vested the Murray Islands in the State to the exclusion of all other rights, interests and claims. The only rights and interests thereby extinguished were those which the Meriam people claimed to hold (and were subsequently found to hold in Mabo v Queensland (No 2) [1992] HCA 23; 175 CLR 1 (Mabo (No 2)) under traditional law as recognised by the common law. Four Justices held that s 10(1) was engaged.

(c)    In the Native Title Act Case, Western Australian statutory provisions that extinguished native title and replaced it with statutory rights were held to engage s 10(1). The new statutory rights had less protection against extinguishment and compulsory acquisition than other property rights. The majority observed that a State law purporting to authorise appropriation of property “characteristically held by persons of a particular race”, for additional purposes or on less stringent conditions than apply to the appropriation of property generally, is inconsistent with s 10(1) (at 437).

(d)    Reference was also made to Ward, where the reasons of Gleeson CJ, Gaudron, Gummow and Hayne JJ referred to the earlier cases concerning native title and s 10(1) (at [117]–[121]). However, Ward was decided after the commencement of the Native Title Act 1993 (Cth) (the NTA) and extensive amendments to that Act in 1998. The issues for decision in Ward concerned the extent to which, under the NTA, certain dealings in land were to be understood as extinguishing native title.

(e)    Maloney concerned a Queensland law that prohibited the possession of alcohol in public places in Palm Island (part of the State of Queensland). The majority held that the law engaged s 10(1), but that it was a special measure within s 8(1) and therefore not inconsistent with the RDA. Maloney was characterised by the Commonwealth as a case about a facially neutral criterion chosen deliberately to achieve an outcome based on race.

2.3.1     Maloney

35    The Commonwealth’s characterisation of Maloney is correct as far as it goes, but ultimately unhelpful.

(a)    It is correct because the criterion applied by the Queensland law was purely geographic: all residents of Palm Island (including a small number of non-Aboriginal persons) were affected by it in the same way; while persons of all races, in places to which the law did not apply, were unaffected by it. Simple comparisons of that kind were rejected as a basis for decision (at [77] (Hayne J), [200] (Bell J)). Instead, members of the majority on the s 10(1) point emphasised that the population of Palm Island was “overwhelmingly” Aboriginal (at [34] (French CJ), [84] (Hayne J, Crennan J agreeing), [362] (Gageler J)), or said that it did not matter that a small number of non-Aboriginal persons was affected by the law in the same way (at [200] (Bell J), [363] (Gageler J)). Bell and Gageler JJ regarded the law as “targeting” an Aboriginal community (at [202] (Bell J), [362] (Gageler J)), while Hayne J described the mischief at which the law was directed as “the evil of alcohol-fuelled violence in [Indigenous] communities” (at [58]). Thus, in our view, the explanation for the majority’s conclusion that the impugned law in Maloney engaged s 10(1) appears to be that their Honours were prepared to look beyond both its text and its immediate operation, and find that geography was being used as a device to impose restrictions directed at Aboriginal persons.

(b)    This understanding of Maloney is nonetheless unhelpful because the authority of a decision of the High Court is not limited to its ratio decidendi. The aspects of the reasoning noted in the previous paragraph show that the impugned law in Maloney was different to the provisions in question in this case, but they are not inconsistent with the provisions in question in this case also engaging s 10(1). The significance of Maloney for present purposes lies not in what it did or did not decide, but in the “seriously considered dicta” of members of the Court concerning the proper understanding of s 10(1) (see Farah Constructions Pty Ltd v Say-Dee Pty Ltd [2007] HCA 22; 230 CLR 89 at [134] (Farah Constructions)).

36    The following aspects of the reasoning in Maloney bear upon the present question.

37    French CJ noted at [11] that s 10 does not apply only to a law that makes a distinction expressly based on race. He said that it is directed to “the discriminatory operation and effect of the legislation” (footnotes omitted).

38    Hayne J (with whom Crennan J agreed) said, at [67]–[68]:

It will be recalled that the RDA is directed to the prohibition and elimination of racial discrimination. These are very general objects and the relevant provisions of the RDA are expressed in very general terms. Section 10 is especially broad. It is directed to the operation of the laws of the Commonwealth and of the States and Territories. It may be contrasted with s 9(1), which makes it unlawful, but not an offence, for a person “to do any act involving a distinction, exclusion, restriction or preference based on race, colour, descent or national or ethnic origin which has the purpose or effect of nullifying or impairing the recognition, enjoyment or exercise, on an equal footing, of any human right or fundamental freedom in the political, economic, social, cultural or any other field of public life”. Whatever the scope of s 9(1), it is sufficient to notice that it contains elements which s 10(1) does not.

In many, perhaps most, cases it will be accurate to describe a law which is found to engage s 10 as a racially discriminatory law. Given the objects of the RDA, that is unsurprising. Care is needed, however, to ensure that this statement of conclusion is not used in a way that inadvertently narrows or confines the operation of s 10. To do so would be contrary to the large objects which the RDA evidently pursues and the generality of the words which it uses … If the law is not a special measure within the meaning of s 8(1), the conclusion that persons of a particular race enjoy a right to a more limited extent than persons of another race is necessary and sufficient to engage s 10.

(Footnotes omitted.)

39    Bell J referred at [200] to earlier decisions in which it had been noted that s 10 does not refer to discrimination or associated concepts, and that it is directed to “the enjoyment of rights by some but not by others, or to a more limited extent by others”. At [201] her Honour said:

Section 10 implements the obligations assumed under Arts 2(1)(c) and 5 of the Convention. In summary, these are the obligations to nullify laws having the effect of creating or perpetuating racial discrimination and to guarantee equality before the law. Equality before the law is the counterpart of the elimination of racial discrimination. Section 10(1) is to be interpreted in the light of these related purposes. A law creates or perpetuates racial discrimination when it applies any distinction, exclusion, restriction or preference based on race which has the purpose or effect of nullifying or impairing the enjoyment or exercise, on an equal footing, of human rights and fundamental freedoms in the political, economic, social, cultural or any other field of public life.

40    At [203]–[204] Bell J dealt with submissions concerning the unintended and anomalous consequences s 10(1) could have if it is understood to apply in cases of unequal enjoyment regardless of whether the law has a legitimate non-discriminatory purpose. The example was given of a planning law requiring buildings in a coastal locality to meet specifications suitable for withstanding extreme weather events. It was suggested that, given a broad reading, s 10 would invalidate that law if the majority of building owners in the relevant location were members of a particular race. Her Honour, having referred to statements in Ward emphasising that s 10(1) is concerned with substance rather than form, suggested that the hypothesised planning law might “not engage s 10(1) because, construed in its context, any limitation on the enjoyment of the right of the building owners would have no connection to race”. No such issue arose in Maloney and it was therefore “not appropriate to determine the extent of the connection with race that is required to validly engage s 10(1)”.

41    Gageler J at [306] described the joint reasons in Ward as having emphasised that:

… s 10 of the RDA is not confined to laws whose purpose can be identified as discriminatory nor to laws that can be said to be aimed at a racial characteristic or to make a distinction based on race and that fulfilment of the condition for the application of s 10 turns rather on the effect of a law on the relative “enjoyment” of a “right” by persons of different races.

(Footnotes omitted.)

42    At [329] his Honour referred to two textual components for the application of s 10. The first was that there exists “a state of affairs in which persons of one race either do not enjoy a human right that is enjoyed by persons of another race or enjoy a human right ‘to a more limited extent’ than persons of another race”. The second was that this was “by reason of” a Commonwealth, State of Territory law. The first component, his Honour said (at [330]), required “no more than that ‘persons’ of one race enjoy a human right ‘to a more limited extent’ than ‘persons’ of another race”.

43    At [331] Gageler J noted that the reference to “persons” in s 10(1) connoted “groups not individuals”. His Honour went on to observe that, nevertheless, it was not necessary for such a group to comprise all members of one or other race. However, the reference to s 10(1)’s concern with groups rather than individuals suggests another point: that questions as to relative “enjoyment” of human rights may in some cases turn on analysis of a group’s circumstances or experience in the realm of the social sciences rather than specific impacts on the enjoyment of rights that can be proved in relation to all members of the group or in relation to the individual applicant.

44    At [333] Gageler J made the following important point:

Persons of one race can enjoy a human right “to a more limited extent” than persons of another race without suffering impairment or infringement of that human right. That proposition can be illustrated by an example adapted from one given by the European Court of Human Rights concerning the requirement of Art 14 of the European Convention that “enjoyment” of the rights and freedoms set forth in that Convention be secured “without discrimination”. A State may well not infringe the human right “to education and training” referred to in Art 5(e)(v) of the Convention by failing to establish a particular kind of educational institution. But if a State establishes an educational institution of a particular kind, the State must ensure that the education the institution provides is available equally to persons of all races. A State law cannot, consistently with s 10 of the RDA, arbitrarily bar the admission of persons of a particular race.

(Footnotes omitted.)

45    His Honour observed at [335] that a difference in the extent of enjoyment of a human right was “a question of degree”, to be answered in the light of the principles and objectives of the Convention.

Construed against the background of those principles and objectives, persons of one race will enjoy a human right “to a more limited extent” than persons of another race where a difference in their relative enjoyment of a human right is of such a degree as to be inconsistent with persons of those two races being afforded equal dignity and respect. The relevant indignity or want of respect lies in the difference in the levels of enjoyment of a human right by persons of the two races rather than in the absolute level of enjoyment by persons of the disadvantaged race. The significance of a difference can be affected by contextual factors, which may include racial targeting or presumptions about the characteristics of racial groups just as they may include ignorance or lack of consideration of the characteristics of racial groups.

46    At [339]–[348] the analysis of Gageler J introduces concepts of proportionality and justification for differential treatment. Some of this reasoning appears to suggest a further condition for the application of s 10(1), namely that the law (so far as it gives rise to different treatment) is not justified by being shown to adopt criteria that are applied in support of a legitimate aim and reasonably necessary to the achievement of that aim (eg at [343]). However, it appears from [347] that this reasoning is looking forward to his Honour’s analysis of special measures under art 1(4) of the Convention and s 8(1) of the RDA.

47    These statements are consistent with the view that s 10(1) is concerned with the practical effect of laws. They are firmly against any limitation on its operation based on whether, in enacting the impugned law, the legislature has set out to make distinctions based on race.

48    That said, it is not clear that the observations of the majority support the position of the applicant in the present case. Care must be taken, of course, because their Honours did not have before them issues of the kind that arise here. For the most part, their observations on s 10(1) are at a fairly high level of generality. The following points suggest that at least some of the majority Justices might not have embraced a submission of the kind being put by the applicant here.

(a)    The hypothetical example of a planning law, discussed by Bell J at [203]–[204], is somewhat closer to the present case. In the example, the law would affect only (or predominantly) members of a particular race because of the composition of the class of persons who owned property in the relevant location. As noted above, her Honour suggested (without elaborating) that, in context, the limitation on the enjoyment of rights might not engage s 10(1) on the basis that it “would have no connection to race”. This suggests that s 10(1) is not necessarily engaged by any law that affects the enjoyment of rights by people of different races in different ways.

(b)    Gageler J at [335] suggested that not every disparity in the extent of enjoyment of a human right would engage s 10(1). The difference would need to be “of such a degree as to be inconsistent with persons of those two races being accorded equal dignity and respect”, and the significance of the difference could be affected by contextual factors. That suggests that, in the case of a law of the kind presently under discussion (conferring an entitlement on the basis of criteria that pay no regard to race), more would need to be shown than that members of one race are less likely to meet its criteria than members of another race.

2.3.2     Intermediate appellate court decisions

49    The Commonwealth also relied on several decisions of intermediate appellate courts. The applicant submitted that, to the extent that these decisions supported the Commonwealth’s position, they were inconsistent with Maloney and should not be followed.

50    We have come to the view that none of these cases is directly on point, so that no question arises as to whether they should be followed. We therefore do not need to express concluded views as to whether any of these cases was (as the applicant submitted) wrongly decided. However, it is useful to say something about two preliminary matters before discussing the cases: the relationship between intermediate appellate decisions and “seriously considered dicta” of the High Court; and the position of a Full Court of this Court constituted (in the original jurisdiction) to determine separate questions.

51    The view has occasionally been expressed that, as a matter of precedent, a decision by a Full Court exercising the original jurisdiction of the Court carries no more weight than three (or five, as the case may be) single judge decisions to the same effect: see the cases cited in Minister for Immigration, Citizenship, Migrant Services and Multicultural Affairs v FAK19 [2021] FCAFC 153; 287 FCR 181 at [31] (Allsop CJ) (FAK19). However, in FAK19 the Chief Justice doubted the correctness of that view, given the text and structure of the Federal Court of Australia Act 1976 (Cth) (FCA Act) and the importance of Full Court authority in the operation of the system it creates: at [32]. The issue did not need to be decided in FAK19. The point did arise, however, in Pitman v Commissioner of Taxation [2021] FCAFC 230; 289 FCR 287 (Pitman), where the Full Court exercising original jurisdiction (in an “appeal” from the Administrative Appeals Tribunal) was invited to depart from a decision of an earlier Full Court exercising the same jurisdiction. Davies J (at [10]) regarded that step as not open unless the earlier decision was “plainly wrong”, and quoted from other parts of Allsop CJ’s reasons in FAK19. Allsop CJ (at [3]) agreed and said:

… for all the reasons set out in FAK19 at [14]–[29], in particular the reasons directed to the text and structure of the [FCA Act], Full Court decisions in the original jurisdiction (often sat for the purpose of expressing a view on a legal or constructional question) should not be viewed somehow as of lesser authority than Full Court decisions in the appellate jurisdiction.

52    The other member of the Court in Pitman, Bromwich J, regarded the earlier decision as clearly correct, but would otherwise not have departed from it for the reasons given by Allsop CJ and Davies J (at [34]).

53    The issue that now arises is different, but related: should a Full Court exercising the original jurisdiction consider itself bound by a decision of an earlier Full Court exercising appellate jurisdiction, or another intermediate appellate court in the integrated judicial system of Australia?

54    Unlike issues concerning the status to be afforded within the Court to existing judgments of the Court, the question is not solely one of practice (cf FAK19 at [24], [32]). It concerns, in part, consistency of approach as between different courts in the Australian judicial system. That is a topic upon which the High Court has expressed views (in Farah Constructions at [135] and earlier in Australian Securities Commission v Marlborough Gold Mines Ltd [1993] HCA 15; 177 CLR 485, 492). However, the position arrived at by the High Court appears to be that even a single judge is not strictly bound by a decision of an intermediate appellate court in another jurisdiction and may depart from it if persuaded that it is plainly wrong (see CAL No 14 Pty Ltd v Motor Accidents Board [2009] HCA 47; 239 CLR 390 at [49]–[50]). The fact that we are now exercising original rather than appellate jurisdiction may therefore not matter. If that is not correct, and a single judge is constrained in this regard to a greater extent than an appellate court, we would nevertheless conclude that a Full Court of this Court exercising original jurisdiction is in essentially the same position as a Full Court exercising appellate jurisdiction. That conclusion is supported by the consideration that under s 24 of the FCA Act a decision of a single judge can be the subject of an appeal to the Full Court, where existing intermediate appellate authority can be canvassed and departed from if it is plainly wrong, whereas a decision of the Full Court (including one exercising original jurisdiction) cannot: the only avenue of appeal in that case is to the High Court, subject to a grant of special leave. Statements of principle in the cases are to be approached on this basis; and the decisions, if not distinguishable, are to be followed unless plainly wrong.

55    An aspect of the “plainly wrong” question in relation to intermediate appellate decisions is whether they can be reconciled with High Court authority. Another way of putting the point is that, as was noted in Hill v Zuda Pty Ltd [2022] HCA 21; 96 ALJR 540 at [25]–[26], an intermediate appellate court should follow “seriously considered dicta” of the High Court; while a decision of another intermediate appellate court on the interpretation of Commonwealth legislation can properly be departed from if it is considered to be “plainly wrong”. In other words, institutional respect between courts of equivalent status gives way, in the event of conflict, to the authoritative status of seriously considered dicta of the High Court. To the extent that a decision of another intermediate appellate court (or this Court) cannot be reconciled with the view of the majority in Maloney concerning the scope of s 10(1), therefore, that decision must be regarded as wrongly decided and therefore not to be followed.

56    With these principles in mind, we turn to the authorities relied on by the Commonwealth. As noted earlier, we have come to the view that these cases are distinguishable and we therefore do not need to express concluded views as to whether they are to be followed, on the one hand, or “plainly wrong” on the other.

57    In Melkman v Commissioner of Taxation (1988) 20 FCR 331 the appellant sought to claim the benefit of a provision of the tax law which exempted a pension paid by a State of the Federal Republic of Germany by way of compensation to victims of Nazi persecution. The appellant was receiving a pension of that character paid by the Netherlands. His attempt to rely on s 10(1) was found at 336–337 to face at least two problems. One was that the human rights referred to in art 5 that he invoked (those found in paras (d)(iii) and (e)(iv)) were considered not to involve exemption from taxation. Another was that the law did not draw any distinction, express or otherwise, between persons of different races. This second point might now require further scrutiny in light of Maloney, because the German pensions to which the exemption applied were payable (in general at least) only to people of German national origin. By exempting those pensions but not others from tax, the exemption provision arguably granted an exemption that was only available to members of a particular race (although there would be a significant issue as to how the relevant “race” was to be identified, for the purpose of working out who were the persons of “another race” entitled to the same benefit by force of s 10(1)). However, this observation does not cast any doubt on the other significant barrier that the appellant’s argument faced: the need to prove that the particular human rights to which he referred were enjoyed to a lesser extent by members of one or more races, as a result of a small number of members of a particular race having access to a tax exemption.

58    Nguyen v Refugee Review Tribunal (1997) 74 FCR 311 (Nguyen) involved a Vietnamese man whose application to the (then) Refugee Review Tribunal (RRT), for review of a decision by an officer not to grant him refugee status, had been filed outside the applicable limitation period. The RRT had no power to grant an extension of time and concluded that it had no jurisdiction to conduct a review. The appellant argued that, because notice of the primary decision had been sent in English, he was less able to enjoy the right to be notified than a person of another race who could understand English, and s 10(1) therefore applied. So far as can be deduced from the report, his argument seems to have been that s 10(1) gave him an entitlement to be notified in his own language; that had not occurred, and therefore time had not started to run.

59    Tamberlin J held that provision of a notice in the de facto official language of Australia could not be said to be discriminatory in form or effect any more than the printing of legislation and judicial decisions in English was discriminatory (at 319). In the alternative, his Honour reasoned that any lessening of the appellant’s enjoyment of the right to be notified arose from his circumstances (ie that he did not understand English) and not as a result of the terms or effect of the legislation. Sundberg J reasoned at 326–327 that the legislation did not give the members of any race a right to receive notice in their own language. Rather, notification was given in English (which, his Honour found, was implicitly required by the legislation) because that was Australia’s official language. Alternatively, if the argument was that the appellant enjoyed the right to “notification” to a lesser extent than a person who could read English, his Honour regarded that argument as unsound because having the notice brought to one’s attention or understanding it was not a component of the requirement for notice under the Migration Act (at 327). Marshall J (at 331) reasoned along similar lines to Sundberg J.

60    The reasoning of Tamberlin J in Nguyen appears to us to be open to question because of its focus on discrimination (a concept s 10(1) does not employ, at least directly) and the suggestion that any lesser enjoyment of a right was the consequence of the appellant’s circumstances rather than the law. The second aspect is problematic for the reason we have indicated above (at [29]-[31]), and also because the circumstance of the appellant that led to his lesser enjoyment (lack of comprehension of English) was, at least prima facie, related to his race. The same problem arises in the present case, where the reduced life expectancy of Aboriginal men compared to their non-Aboriginal contemporaries is expressly agreed to be a “function” of their race. If lesser enjoyment of a human right by members of the Aboriginal race could be said not to arise “by reason of” the law that gives local expression to that right, on the footing that lesser enjoyment is a consequence of factors arising from their Aboriginality, s 10(1) would be readily circumvented. However, these criticisms do not affect the reasoning of Sundberg and Marshall JJ or the correctness of the actual decision. The practice (widespread if not universal among nation states) of conducting government business in one or two official languages clearly raises special issues, which do not arise in this case. The proper understanding of the relationship between an official language and art 5 of the Convention (and with s 10(1)) may be found in the proper identification of the relevant human right (a point that received little attention in Nguyen) and close attention to what is meant by the enjoyment — and the relative enjoyment — of that human right. Nguyen has not been shown to be wrongly decided, but we do not think it assists greatly in resolving the present case.

61    Sahak concerned applications for judicial review of decisions by the RRT, lodged out of time by applicants who were unable to read and write in English. The focus was s 478 of the Migration Act, which imposed a time limit for an application invoking the statutory review jurisdiction of the Federal Court under s 476. The appellants had both filed review applications outside the permitted period, which the Court had no power to extend. They argued that, as a result of s 478, persons of Syrian and Afghan national origin enjoyed the right to equality before the law, including access to the Court, to a more limited extent than persons of a national origin which had an attribute or characteristic of English as a first language; and that therefore, by force of s 10(1), the time limit did not apply to them. Goldberg and Hely JJ observed at [45] that the discrimination or disadvantage arising from the practical operation of the time limit was not racial discrimination, because a person whose national origin was Afghan or Syrian was able to take advantage of the relevant right to the same extent as anybody else if they had a sufficient understanding of English or if they had access to friends or professional interpreters.

62    Their Honours then said, at [48]:

The fact that an applicant who wishes to review the decision of a Tribunal requires the services of an interpreter in order to prepare and file an application for review does not mean that the right to apply for the review is lessened. Similarly, a person who speaks English but who does not understand how to complete the application due to circumstances, such as physical infirmity, a lack of literacy or a lack of education, does not have his or her right to apply for review lessened by the time limit in s 478 compared to the right of a literate, educated, healthy, English-speaking applicant. Any difficulty such persons confront in completing and filing applications for review within the time limit prescribed by s 478 is due to personal characteristics and not due to a circumstance which is dictated by their race, colour, or national or ethnic origin.

63    Their Honours went on at [49] to conclude that there was no nexus or causal connection between the provisions of s 478 and the manner in which the appellants enjoyed their right of access to the Court as compared with the manner in which English-speaking applicants enjoyed that right. Section 478 operated uniformly and, in their Honours’ view, did not have a differential or discriminatory impact.

64    We have expressed doubts about this last aspect of the reasoning above. Part 8 of the Migration Act, as in force at the relevant time, created a right to seek judicial review of certain migration decisions in the Federal Court in addition to the constitutionally entrenched jurisdiction of the High Court under s 75 of the Constitution. Part of the manner in which that right was conferred was that it could only be exercised by making an application within a strict time limit. To the extent that the right was enjoyed to a lesser extent by members of one race compared to members of another, and that lesser enjoyment arose from the more onerous application of the time limit upon people whose first language was not English, it is hard to see how that lesser enjoyment did not have a nexus or causal connection with the provisions of s 478.

65    However, the result in Sahak can be explained on the basis of what their Honours said at [48], which echoes the reasoning in Nguyen. It can also (and perhaps more persuasively) be explained on the alternative basis set out in the reasons of North J (at [3]–[4]). If the relevant human right is understood as the right to seek judicial review of an adverse decision, and regard is had to the jurisdiction of the High Court under s 75(v) of the Constitution (which, unlike the jurisdiction of the Federal Court under Pt 8 of the Migration Act, was not limited to specified grounds), it could not be said that the right was not enjoyed or enjoyed to a lesser extent by the appellants than by other persons.

66    Jones v Public Trustee of Qld [2004] QCA 269; 209 ALR 106 (Jones) concerned provisions in the Succession Act 1981 (Qld) (Succession Act) specifying the persons who could pursue proceedings in the Supreme Court concerning an estate that had been administered on intestacy. The appellant claimed the right to pursue such proceedings on the basis that he was the most senior elder of the Dalungdalee people, to which the deceased had belonged. That was not a sufficient interest under the provisions of the Succession Act. It was argued that the relevant provisions were inconsistent with s 10(1), on the basis that they conflicted with Aboriginal traditional or customary law. However, McPherson JA (with whom Williams and Jerrard JJA agreed) held that the content of any such traditional rights had not been proved. His Honour also observed (at [19]) that the provisions of the Succession Act made no distinction between peoples of any race or origin and applied equally to all people including Aboriginal people. In the light of Maloney, the latter proposition might well not be sufficient to dispose of the case. However, it may be doubted whether “the right to inherit” (referred to in art 5(d) of the Convention) is enjoyed to a lesser extent by the people of any race in a situation where the same rules apply in relation to any person who dies intestate. The case is, rather, one where the State has chosen to apply the same rules to everyone rather than create different intestacy regimes for people of different racial or cultural backgrounds. In the end, we do not think Jones assists in the resolution of the present issues.

67    In Vanstone v Clark [2005] FCAFC 189; 147 FCR 299 (Vanstone), the respondent was the Chair of the Aboriginal and Torres Strait Islander Commission (ATSIC). He sought judicial review of a decision by the appellant (as the responsible Minister) to suspend him from office under a provision of the Aboriginal and Torres Strait Islander Commission Act 1989 (Cth) (ATSIC Act). The decision depended on a determination made (purportedly) under another section of the ATSIC Act, providing that conviction for certain offences was to be taken to constitute “misbehaviour” for the purposes of that Act. At first instance, the determination (and thus the decision) was held invalid on several grounds, one of which was that its effect was to hold office-holders of ATSIC to a higher standard than holders of comparable offices under other legislation. Section 10 was said to require the determination to be read down to overcome that effect (Clark v Vanstone [2004] FCA 1105; 211 ALR 412 at [109]–[115]). In the Full Court, the other statutory offices referred to by the primary judge were regarded as not comparable (Vanstone at [201] (Weinberg J (Black CJ agreeing))), and the determination did not operate in a discriminatory way between Indigenous and non-Indigenous holders of offices under the ATSIC Act (at [200]). There was thus no inconsistency of treatment based upon race.

68    Queensland Construction Materials Pty Ltd v Redland City Council [2010] QCA 182; 271 ALR 624 (Queensland Construction Materials) concerned a provision in Queensland planning law entitling the owners of neighbouring properties to be given notice of, and object to, a proposed development. “Owner” was relevantly defined to include a person entitled to receive rent from the property (or who would be so entitled if the property were leased). Parties who claimed to hold native title in certain land contended that, by virtue of s 10(1), they were entitled to be given notice. Chesterman JA and Applegarth J held that, even if native title was established, s 10(1) would not operate so as to put the claimants in the same position as holders of other interests in land. Their Honours held that the distinction made by the relevant provisions was based not on race but upon different proprietary interests in land. If the claimants were able to establish a right to exclusive possession and a right to receive rent for the land, they would relevantly be “owners”; otherwise, along with holders of various kinds of non-native title rights other than freehold and leasehold, they would not be.

69    An application for special leave to appeal to the High Court in Queensland Construction Materials was refused on the papers: [2011] HCASL 131. It is not clear whether the issue referred to in the previous paragraph was raised in the application.

70    In R v Woods [2010] NTSC 69; 246 FLR 4 (Woods), the defendants were two Aboriginal men charged with murder. They challenged the array of jurors on several bases, including that provisions in the Northern Territory legislation excluded persons who had been in custody in the previous seven years from jury service and required the service of summons by ordinary post. These provisions were said to result in a disproportionately low number of Aboriginal people on jury panels, because of their over-representation in the prison population and because many had no postal address. This, it was argued, engaged s 10(1) because the defendants, compared to non-Aboriginal persons, had less chance of being tried by a jury including people of the same race as them. The argument was rejected by the Full Court, holding that the right to trial by jury did not include any requirement as to the racial mix of the jury panel. Rather, in their Honours’ view, random selection of jury members was an important element of trial by jury (at [58]). The Court went on to say, at [59]:

To impose some overriding requirement to the effect that a jury, once randomly selected in this way, has to be racially balanced or proportionate would be the antithesis of an impartially selected jury, not to mention the enormous practical difficulties that would be associated with attempting to meet such a requirement, particularly as it is not an easy matter to identify who is, or is not, a member of a particular racial group.

71    Factually, Woods is close to the present case. The impugned law was part of the statutory regime giving effect under municipal law to one of the fundamental human rights whose equal enjoyment is mandated by the Convention. Its facially neutral criteria were said to operate differently as between Aboriginal people and others, because of social facts concerning the life experiences of Aboriginal people that resulted in different probabilities as to how relevant rights would be enjoyed. The resolution of the case lay in analysing the relevant human right and what was involved in its equal or unequal enjoyment. In our view, the reasoning in Woods is not overtaken by anything that was said in Maloney.

72    Aurukun is a case that might well be approached differently following Maloney, although that does not mean it was wrongly decided. It concerned amendments to the Liquor Act 1992 (Qld) (Liquor Act) which provided that a local government entity could no longer hold a liquor licence. The only local government entities that held liquor licences were Indigenous local governments.

73    Keane JA held that s 10(1) was not engaged on the basis that neither the right to use a local government facility to purchase alcohol nor the right to use a local supplier was a human right or fundamental freedom of the kind with which s 10 was concerned (at [155]); the provisions were not in breach of s 10 because they served a legitimate and non-discriminatory goal (protecting women and children in Indigenous communities from domestic violence) (at [166]–[170]); and the law did not result in a different level of enjoyment of the right to have access to a local supplier of alcohol (the practical effect of the law being that nobody in Queensland could acquire alcohol from their local government) (at [178]–[179]). Alternatively, his Honour would have held that the law was a special measure within s 8 of the RDA. Philippides J held, as to the rights of individuals, that the provisions were not inconsistent with equal treatment before the law because they applied everywhere in Queensland (at [262]) and other means of obtaining alcohol were not precluded (at [275]–[276]); and that, so far as there was interference with the property rights of the local councils, those rights were not absolute and could be modified to achieve a legitimate and non-discriminatory public purpose (at [266]–[271]). McMurdo P held that s 10(1) was engaged, but that the provisions were a special measure.

74    An application for special leave to appeal from the decision in Aurukun was dismissed on the basis that an appeal would have insufficient prospects of success. Refusal of special leave has no status as precedent (North Ganalanja Aboriginal Corporation v Queensland [1996] HCA 2; 185 CLR 595, 643 (McHugh J)), although in some circumstances the reasons given can be persuasive (Algama v Minister for Immigration [2001] FCA 1884; 115 FCR 253 at [62] (Whitlam and Katz JJ, French J agreeing)). The short oral reasons delivered by Hayne J indicate that the High Court was not persuaded that the applicants had significant prospects of identifying a fundamental right or freedom that was infringed (Transcript of Proceedings, Aurukun Shire Council v CEO, Liquor Gaming & Racing in Dept of Treasury [2010] HCATrans 293), and therefore do not involve any endorsement of the reasoning in the Court of Appeal concerning the practical effect of the provisions. (The provisions considered in Aurukun, while they amended the Liquor Act, were part of an Act called the Aboriginal and Torres Strait Islander Communities (Justice Land, and Other Matters) and Other Acts Amendment Act 2008 (Qld) which was plainly directed at Indigenous communities. There is every reason to think that, if these provisions had affected the enjoyment of a fundamental human right of the kind referred to in art 5 of the Convention, the Court as constituted in Maloney would have regarded them as engaging s 10(1); although the provisions might well also have been regarded as a special measure.)

75    Munkara is the first of the intermediate appellate decisions to which we were referred that post-dates Maloney. Section 6 of the Alcohol Protection Orders Act 2013 (NT) (APO Act) provided police with powers to issue alcohol protection orders. These orders could be issued if a person was charged with a “qualifying offence” and the issuing officer believed that the person was affected by alcohol at the time. The effect of an order was to prohibit the person from possessing or consuming alcohol or (with certain exceptions) being on licensed premises. Evidence showed that Aboriginal people were vastly more likely to be issued with alcohol protection orders than other people in the Territory. The appellant argued that s 6 engaged s 10(1) of the RDA and was therefore invalid. He also argued that the provisions for seeking reconsideration of an order engaged s 10(1) because the very short time limits in ss 9 and 11, together with other formalities, created particular difficulties for Aboriginal people given their low levels of literacy in English. The time limits were said to be of no effect for this reason. Although the statute was racially neutral in its terms, the appellant argued (relying on Maloney) that its “legal and practical operation” was to disadvantage Aboriginal people in their enjoyment of the right to own property (ie alcohol) and their right of access to licensed premises (including hotels, restaurants, cafes and some supermarkets).

76    The reasoning of the Court of Appeal on the s 10 issues is to be found in the reasons of Blokland J. Kelly J agreed with that reasoning (at [15]) and Barr J agreed with Kelly J on all of the issues in the appeal.

77    In relation to s 6, Blokland J said at [99]:

The primary judge was correct in holding that any adverse effect suffered by Aboriginal persons as a result of the imposition of an alcohol protection order is not as a result of the law itself but as a result of the person committing a qualifying offence whilst affected by alcohol. The situation is clearly distinguishable from the circumstances considered in Maloney because in that case, the Queensland legislative and regulatory scheme was directed at the largely Aboriginal population of one community, Palm Island, the residents of which were almost all Indigenous persons. They suffered disadvantage without any wrongdoing or qualifying conduct on their part. Under the Act the subject of this appeal, the law does not have effect unless and until a person commits a qualifying offence and a police officer reasonably believes that a person was affected by alcohol at the time of offending. Criminal offending in circumstances where the offender is affected by alcohol triggers the operation of the Act.

78    Her Honour’s response to the argument that alcohol consumption (and therefore exposure to orders under the APO Act) was causally related to race was as follows (at [102]):

That contention is rejected. It adds nothing to the bare statement of the statistics, i.e. that Aboriginal people form about 27 percent of the population and have received over 90 percent of the alcohol protection orders. It could also be said that “serious domestic violence and race are interrelated concepts when it comes to Aboriginal people in the Northern Territory” because the statistics are much the same. Put that way, the concept is simplistic and offensive. The appellant’s submission ignores the reality of deprivation and disadvantage that are the real matters referred to by the High Court in [Bugmy v The Queen [2013] HCA 37; 249 CLR 571] and are well acknowledged to be important factors associated with criminal behaviour and alcohol and substance abuse. Further, it ignores the fact that most Aboriginal people do not abuse alcohol and that many non-Aboriginal people do.

79    However, what her Honour described as the “main point” (at [103]–[104]) was that, unlike the law considered in Maloney, the APO Act did not deprive anyone of the right to possess alcohol or enter licensed premises. Instead, its effect was “to place consequences on people’s behaviour, namely in the first instance committing a qualifying offence while affected by alcohol”. If s 10(1) was engaged because members of a particular race were disproportionately likely to engage in the relevant behaviour, the same would be true of any offence-creating provision in the Criminal Code where members of a particular race are more likely than other people to be imprisoned (and thereby deprived of liberty) for committing the relevant offence.

80    As to the challenge to the time limits on reconsideration requests, Blokland J held that the reasoning in Sahak was not contrary to established authority concerning s 10(1) (including Maloney) (at [118]) and not clearly wrong (at [119]).

81    While we have reservations as to whether some of the reasoning on s 6 of the APO Act in Munkara is consistent with statements in Maloney, it is clear that the Court considered the effect of that case. While the considered dicta in Maloney are binding (to the extent that a majority position emerges from them), comity between intermediate appellate courts requires a high degree of respect to the understanding of the Court of Appeal concerning what flows from those dicta. Thus, in order to conclude that Munkara was wrongly decided by reason of inconsistency with Maloney, we would need to be persuaded that the Court was clearly wrong in its understanding of the effect of that case. In any event, what Blokland J described as the “main point” appears to us, with respect, to be sound. A law of general application imposing penal consequences for behaviour considered to be antisocial would not normally be said to result in a lesser enjoyment of human rights by members of a particular race, unless the behaviour was singled out for punishment as an attack on members of a particular race who were associated with it. (We note here that we have some difficulty with the alternative explanation suggested by Basten JA in Hamzy v Commissioner of Corrective Services NSW [2022] NSWCA 16; 400 ALR 507 at [33] (Hamzy) as to why a criminal law of general application could survive s 10(1) despite having a disproportionate operation on Aboriginal communities: ie that such a law was reasonable and justifiable because it is protective of the same communities. Section 10(1) does not call for any inquiry into whether laws are reasonable or justifiable: Maloney at [39] (French CJ), [68] (Hayne J (Crennan J agreeing at [112])), [167] (Kiefel J), [214] (Bell J), [348] (Gageler J). Nor does s 10(1) invite consideration of whether a detriment that falls disproportionately on some members of a particular race (eg those who engage in criminalised behaviour) is offset by a benefit to other members of the same race (victims of that behaviour).)

82    The reasoning in Munkara on the time limits for seeking reconsideration in ss 9 and 11 of the APO Act grapples directly with whether Sahak is still good law following Maloney. For present purposes we do not need to express a concluded view on that question, as the facts in Sahak are far removed from those in the present case. We have expressed reservations about the majority reasoning in that case. We add that, for the reason outlined in the previous paragraph, in order to regard Sahak as having been overtaken by Maloney, we would need to be persuaded that the Court in Munkara had erred on this issue.

83    The appellant in Hamzy was a prisoner, who had been designated as an “extreme high risk restricted inmate”. As such he was subject to various measures including a requirement that most of his communications be in English (so that corrections officers, who monitored his communications, could know what he was discussing). That placed a burden on him because he was accustomed to speak partly in Arabic, his parents did not speak good English, and speaking in Arabic was important to the practice of his Islamic faith. He argued, among other things, that the requirement to use English was inconsistent with s 10(1).

84    Leeming JA, with whom Bathurst CJ agreed, dismissed the s 10(1) claim on the basis that there was “no other law which conferred a right enjoyed by persons of another race, colour or national or ethnic origin which, by dint of s 10, could be relied upon by the appellant” (at [274]). Basten JA also appears to have embraced that reasoning (at [21]). We doubt, with respect, whether that was an adequate answer to the claim. If the freedom to speak to family members or practice a religion in one’s “native” language (or chosen language) is a human right of the kind protected by art 5 of the Convention, persons in Australia generally enjoy that human right because the common law permits that which is not prohibited (see, eg, Lange v Australian Broadcasting Corporation [1997] HCA 25; 189 CLR 520 at 564 (the Court)). It is not clear to us why a particular law giving effect to the right needs to exist, in order for a law curtailing its enjoyment by members of a particular race to engage s 10(1). However, that point does not need to be pursued here. The human right in question in the present case has been given domestic legal force by statute, and it is the manner in which that has occurred that is in question.

85    Basten JA, in extensive obiter dicta, criticised reasoning of the primary judge to the effect that there would be no breach of s 10(1) “if a person does not enjoy a human right, or does so to a lesser extent, because of his or her individual personal circumstances”. His Honour said that such reasoning could not be accepted; and, to the extent that it was supported by reasoning in Sahak, that reasoning should not be followed (at [23]–[24]). His Honour was critical of some of the reasoning in Sahak (at [25]–[29]) and in Munkara (at [30]–[33]). After referring to Maloney, his Honour went on to state (at [39]–[40]):

To attribute the effects of the clauses of the Regulation impugned by the appellant to “the personal circumstances of his being in custody”, would be analogous to dismissing Ms Maloney’s claim on the basis that the adverse operation of the law was attributable to her residing on Palm Island. Such an approach cannot be sustained. Even if the judge’s reason were an elliptical reference to the restrictions being justified by the circumstances of the appellant’s incarceration, that would involve a rejection of the core of the reasoning as to justification adopted in Maloney.

Further, if that basis of the judge’s reasoning as to the non-engagement of s 10 were applied, it would seem that any racially discriminatory treatment of prisoners would be beyond the protection of s 10. But that would contradict the uniform approach of international bodies to such cases.

86    After discussing some decisions of the European Court of Human Rights on this point, Basten JA turned to the analysis of s 9 of the RDA, which was also raised in Hamzy. In that context, at [69]–[88], his Honour discussed the identification of the relevant right or freedom. At [89]–[91] his Honour concluded that, in the light of the way the case was run at trial (including the ways that the relevant right had been framed), the lawfulness of the challenged provisions could not be determined and the primary judge had been correct to dismiss the summons.

87    The most recent intermediate appellate decision that needs to be mentioned is Bara. The Misuse of Drugs Act 1990 (NT) contained two offences of trafficking less than a commercial quantity of a Schedule 2 drug. Section 5A applied generally in the Northern Territory and prescribed a maximum penalty of five years’ imprisonment. Section 5D applied where the drug was supplied in an Indigenous community, and prescribed a maximum penalty of nine years’ imprisonment. A charge under s 5D also carried a presumption against bail. The appellant, an Aboriginal man, was charged with an offence under s 5D. He argued that the section was invalid by reason of inconsistency with s 10(1).

88    The Court of Appeal held that s 5D was a special measure for the purposes of s 8 of the RDA, and therefore not subject to Pt II of that Act. However, the Court also considered whether, apart from s 8, the impugned provision engaged s 10(1). Facts were agreed between the parties, including that around 25 percent of the Territory’s population were Indigenous and about 87 percent of the population of “indigenous communities” (as defined for the purposes of s 5D) were Indigenous. The agreed facts also included statistical information which showed that 94.9 percent of the charges under s 5D, and 93.6 percent of charges leading to conviction in the Local Court, were against Indigenous people, compared with 49.4 percent of the total charges and 50.7 percent of charges leading to conviction under s 5A; and that sentences under s 5D were significantly more likely to include terms of imprisonment. In the period covered by these figures, as the primary judge observed, 140 Indigenous people were charged and either convicted or committed to the Supreme Court under one or other of the provisions (compared to 60 non-Indigenous people). Forty-seven of these were convicted or committed under s 5A (compared to 52 non-Indigenous people) and 93 (compared to 8) under s 5D. Digging a little deeper, more people were committed to the Supreme Court for offences under s 5A (indicating more serious offending) than s 5D. The primary judge observed that those imprisoned were “by no means overwhelmingly” offenders who had been charged under s 5D.

89    The rights that were said to be unequally enjoyed as a consequence of s 5D were the right to liberty and the right to equal treatment before courts and tribunals. Both were accepted as falling within s 10(1). As to the former, the appellant’s case boiled down to an argument that s 5D caused Aboriginal people to enjoy the right to liberty to a more limited extent than others in the Territory because it extended the time that they spent in custody, compared with non-Aboriginal people charged under s 5A (at [40]). As to the latter, the Court referred at [44] to statements in Maloney that were critical of proposing a broadly expressed right to equality as a right that engages s 10(1), and regarded the right referred to in art 5(a) of the Convention as concerned essentially with procedural equality. Debate on this issue thus focused on whether the presumption against bail, and a mandatory minimum sentence that applied in some circumstances, were “procedural” or “substantive” in character: at [45].

90    The appellant’s argument on unequal enjoyment of rights was understood to involve a comparison between people charged under s 5D (who were predominantly Aboriginal) and people charged under s 5A. The argument invoked reasoning in Maloney (where, it will be recalled, the population of Palm Island was overwhelmingly but not entirely Aboriginal). The Court held that this division did not involve a meaningful comparison, because almost half of the people charged under s 5A were also Aboriginal (at [48]). A further difficulty identified by the Court was that the comparison sought to be made was between people charged with different kinds of offending: offences under s 5D were more serious because of the aggravating factor of supplying drugs to a vulnerable community (at [49]). Thus, the argument was to be rejected.

91    The Court went on to identify further problems that it perceived in the appellant’s argument.

(a)    The case was said to be different from Maloney because s 5D did not make anything unlawful that was previously lawful. Section 5D had no impact on any person’s rights or freedoms unless and until the person engaged in or was reasonably suspected of engaging in criminal conduct (matters which were “not related to race”) (at [57]). Munkara was a similar case, and its reasoning was therefore not inconsistent with Maloney (at [58]–[60]).

(b)    In response to the point that the reference to Indigenous communities in s 5D was a criterion based on race (and its disproportionate application was therefore not incidental or accidental), the Court observed that the facts before it did not show what proportion of the Aboriginal people charged with the offence actually lived in or had connections with those communities (at [63]). Experience showed that, in general, Aboriginal people were more likely to be involved at lower levels of organised drug supply, making them more visible and liable to be caught (at [64]). It was therefore “difficult to draw conclusions about the strength of the causal nexus between the number of Aboriginal people charged with an offence against s 5D and the reference in s 5D to indigenous communities” (at [65]).

(c)    In any event, connection with an Indigenous community did not of itself give rise to any liability. The necessary condition for liability was engagement in the supply of dangerous drugs (referring again to Munkara). Section 5D thus did not turn on a criterion based on race (at [66]–[67]).

(d)    Exposure to a higher maximum penalty, compared with offences under s 5A, did not result in a lesser enjoyment of human rights because the maximum penalty was only one, usually relatively minor, driver of the ultimate sentencing disposition (at [73]–[76]) and it had previously been part of the sentencing practice of courts in the Territory to treat supply to an Indigenous community as an aggravating factor in cases of the supply of dangerous drugs (at [77]–[78]). A person charged under s 5D and convicted would most likely receive a very similar penalty to someone who had engaged in identical conduct and been charged under s 5A (at [79]–[83]).

(e)    The mandatory minimum penalty (28 days’ imprisonment) applied to an offence for which the maximum penalty was seven years’ imprisonment or more, and therefore to offences under s 5D, unless the sentencing court was of the opinion that such a penalty should not be imposed. It therefore did no more than reverse the normal approach of imposing a custodial sentence as a last resort. The mandatory minimum also applied to other offences (including under s 5A) if an “aggravating circumstance” was present. “Aggravating circumstance” was defined to include an offence committed in an Indigenous community. Thus, the mandatory minimum penalty did not lead to any significant disparity in enjoyment of the right to liberty between offenders charged for the same conduct between ss 5A and 5D (at [84]–[89]). For the same reason there was no difference in enjoyment of the right to equal treatment (at [103]).

(f)    Analysis of the statistics showed only a relatively modest difference in imprisonment rates between offenders sentenced for offences against the two provisions. In a relatively small sample, the difference was insignificant and explicable by differences in the seriousness of offending or the subjective circumstances of offenders (at [90]–[92]).

(g)    In the absence of evidence it was unsafe to assume that the presumption against bail for people charged with offences under s 5D led to Aboriginal people charged with such offences being refused bail in significantly greater numbers than non-Indigenous people charged under s 5A (at [98]–[101]).

(h)    The provisions imposing the presumption against bail and the mandatory minimum sentence were matters of substantive law, not procedure, and therefore did not lead to any inequality of treatment by courts or tribunals (at [102]–[105]).

2.3.3     Conclusions on the case law

92    We do not consider that any of the intermediate appellate decisions discussed in this part of the reasons needs to be regarded as wrongly decided in the light of Maloney. However, the statements in Maloney referred to earlier and the statements by Basten JA in Hamzy lead us to make the following observations.

93    First, there is a danger in expressing reasoning on s 10 in terms suggesting that lesser enjoyment of a human right arises from a person’s “personal circumstances” or “personal characteristics”. A person’s personal circumstances or characteristics include their racial background, so that, without some further explanation, the terminology does not help to explain why s 10(1) has not been engaged. Indeed, there is a risk that a differential impact which ought to be seen as engaging s 10(1) could be defined out of existence by labelling it as an aspect of “personal circumstances”. Thus, it was no answer to the appellant’s case in Maloney to say that she was affected by the impugned law because of her personal circumstance of living on Palm Island.

94    Language of this kind appears in Sahak at [48], which we have set out above. The proposition that the appellants in that case were hindered in complying with the time limit for lodging review applications by their “personal characteristics” is difficult to accept as an explanation for why s 10(1) was not engaged. The suggestion (also evident at [45]) that inability to understand and respond in written English was not a characteristic linked to race (because English speakers who are uneducated or physically infirm face the same problem) is also difficult to accept, quite apart from what was said in Maloney (cf Hamzy at [26]). As noted above, we do not need to express a concluded view as to whether Sahak was correctly decided; and the decision may be explicable on a basis that does not depend on this reasoning. (We note that the explanation suggested by Basten JA in Hamzy at [80] — that the correct approach is “not to deny the unequal treatment based on race, or national or ethnic origin, but to accept that the adoption of a single language of public administration is reasonable and justifiable, so long as any disadvantage is ameliorated for example by the provision of interpreters” — does not work for s 10(1) because the provision makes no exceptions for laws that are reasonable and justifiable.)

95    The laws considered in the recent Northern Territory cases of Munkara and Bara imposed consequences on people by reference to their conduct or suspected conduct. However, there was a persuasive case that such conduct was (at least statistically) strongly linked to race; and thus there was (prima facie) an unequal impact, as between members of different races, on the likelihood of their being deprived of liberty. The fact that penal consequences arose from personal conduct is therefore, in our view, not sufficient to explain why s 10(1) was not engaged. The preferable way to understand these cases is that they rest on the proposition that a generally applicable criminal law is (at least normally) not inconsistent with the equal enjoyment of human rights even if it affects people of different races to different extents. The outcome would likely be different if the conduct singled out for punishment were characteristic of members of a particular race.

96    Secondly, as Gageler J noted in Maloney at [331], s 10(1) applies by reference to groups rather than individuals. It is therefore neither a necessary nor a sufficient condition for its engagement that an individual applicant be adversely affected in their enjoyment of human rights. (The applicant in Nguyen, for example, was not adversely affected in any relevant way by anything arising from his racial background — his review application had been sent to the Tribunal well before the relevant deadline and was held up in the post — but this was not suggested as an answer to his claims.) What must be shown is that a law has an unequal application (either in its terms or in its practical effect) as between members of a race (or a subset of them), taken as a group, and members of another race. That involves unequal application or effect either directly by reference to membership of that race or by reference to some attribute linked with or shared by members of that race and not others (eg possession of native title rights).

97    Hence, although arguments framed by reference to statistics were found unpersuasive in Woods, Munkara and Bara, there is no reason in principle why the unequal enjoyment of a human right, as between members of different races, cannot be established by reference to probabilities or likelihoods that are proved by analysis drawn from the social sciences. Any case involving a law that employs facially neutral criteria is likely to involve analysis of that kind. In Maloney, the crucial factor was that most (not all) residents of Palm Island were Aboriginal people. It is therefore not necessarily fatal to the applicant’s case that the agreed facts establish life expectancies (ie anticipated average lifespans based on analysis of populations) rather than the period for which he or any other person is going to live. Facts of that kind are capable of establishing an attribute or characteristic, associated with membership of a race, whose interaction with the impugned law may engage s 10(1). However, there are always limits to how far statistics go and what they can be used to prove. In Fraser v Canada (Attorney-General) [2020] SCC 28, a case concerning art 15 of the Canadian Charter of Rights and Freedoms, Abella J (writing for the majority) said (at [56]–[60], citations omitted):

Two types of evidence will be especially helpful in proving that a law has a disproportionate impact on members of a protected group. The first is evidence about the situation of the claimant group. The second is evidence about the results of the law.

Courts will benefit from evidence about the physical, social, cultural or other barriers which provide the “full context of the claimant group’s situation”. This evidence may come from the claimant, from expert witnesses, or through judicial notice. The goal of such evidence is to show that membership in the claimant group is associated with certain characteristics that have disadvantaged members of the group, such as an inability to work on Saturdays or lower aerobic capacity. These links may reveal that seemingly neutral policies are “designed well for some and not for others”. When evaluating evidence about the group, courts should be mindful of the fact that issues which predominantly affect certain populations may be under-documented. These claimants may have to rely more heavily on their own evidence or evidence from other members of their group, rather than on government reports, academic studies or expert testimony.

Courts will also benefit from evidence about the outcomes that the impugned law or policy (or a substantially similar one) has produced in practice. Evidence about the “results of a system” may provide concrete proof that members of protected groups are being disproportionately impacted. This evidence may include statistics, especially if the pool of people adversely affected by a criterion or standard includes both members of a protected group and members of more advantaged groups.

There is no universal measure for what level of statistical disparity is necessary to demonstrate that there is a disproportionate impact, and the Court should not, in my view, craft rigid rules on this issue. The goal of statistical evidence, ultimately, is to establish “a disparate pattern of exclusion or harm that is statistically significant and not simply the result of chance”. The weight given to statistics will depend on, among other things, their quality and methodology.

Ideally, claims of adverse effects discrimination should be supported by evidence about the circumstances of the claimant group and about the results produced by the challenged law. Evidence about the claimant group’s situation, on its own, may amount to merely a “web of instinct” if too far removed from the situation in the actual workplace, community or institution subject to the discrimination claim. Evidence of statistical disparity, on its own, may have significant shortcomings that leave open the possibility of unreliable results. The weaknesses with each type of evidence can be overcome if they are both present. Prof. Colleen Sheppard (2001) recognizes this possibility:

While in some cases the overwhelming correspondence between certain categories and the gender or racial composition of the category makes the sex or race discrimination claims relatively easy to substantiate, in other cases the statistical preponderance may be less marked. In such cases it may also be important to consider the qualitative components of the harm that constitutes discrimination.

98    Thirdly, the statements in Maloney referred to above prevent acceptance of the Commonwealth’s explanation for the decisions in the intermediate appellate court cases: that is, that s 10(1) is engaged only by a law that turns on an expressly race-based criterion or that is deliberately aimed at people of a particular race. Those statements, and the text of s 10(1) itself (especially when read with art 5 of the Convention), make it clear that the provision is concerned at least in part with the practical effect of laws: whether the law results in the unequal enjoyment of a human right or freedom, of the kind referred to in art 5, as between the people of different races.

99    Fourthly, rather than a limit on the reach of s 10(1) of the kind proposed by the Commonwealth, the point that emerges is that not every differential impact involves the unequal enjoyment of human rights. Woods illustrates the point. Aboriginal people charged with indictable offences were shown to have a significantly reduced chance, compared to other accused persons, of being tried by a jury of people of their own cultural background. However, the right to trial by jury was analysed and found not to involve a requirement for racial representativeness.

100    Two aspects of the reasoning in Maloney, referred to at [45] above, are relevant here. First, at [335] Gageler J described the unequal enjoyment of a human right for the purposes of s 10(1) as a question of degree, potentially affected by contextual factors, and arising where a difference in the enjoyment of a human right is “of such a degree as to be inconsistent with persons of those two races being afforded equal dignity and respect”. Secondly, questions of context, degree, dignity and respect may well have a bearing on the treatment of the hypothetical planning law mentioned by Bell J at [203]–[204]. The effect of such a law on the enjoyment of the right to own property, considered as a fundamental human right, might take on a different aspect depending on the nature of the buildings concerned and the relationship of their owners with the locality. In this sense, and perhaps others, any differential impact of the law might in some circumstances have “no connection to race”.

101    Fifthly, identification of the relevant human right is likely to be important. So too is identification of how the impugned law is said to result in the unequal enjoyment of that right, in the sense discussed above. There must be some interaction between the operation of the law and membership of a race, or a characteristic or attribute associated with membership of a race, that leads to unequal enjoyment of the right.

102    Sixthly, consideration of unequal enjoyment of a human right — such rights not being limited to rights that are legally enforceable under municipal law — may involve complex and debatable evaluative judgments. In a case such as the present, where the effect of s 10(1) if engaged is to augment the rights conferred by municipal law, determining what is the effect of s 10(1) may be particularly complicated. If what is required to be overcome by a declaration as to the effect of s 10(1) is the unequal enjoyment of a human right flowing from the practical effect of a law, there is some danger that the court may find itself beyond the region of judicially discoverable and manageable standards (to borrow language from Baker v Carr (1962) 369 US 186 at 217 (Brennan J) and Buttes Gas and Oil Co v Hammer [1982] AC 888 at 938 (Lord Wilberforce)). Questions of remedy are not currently before us; and the fear of disorientation is not a sufficient reason to ignore the force of the statements in Maloney to which we have referred. However, it needs to be kept in mind in the construction of s 10(1) that the unequal enjoyment of human rights to which it refers, and the connection of that inequality to race, must be (first) identifiable by making findings on the basis of admissible evidence, and (secondly) remediable by the grant of some relief identifying how, by force of s 10(1), equality of enjoyment of the human right is brought about.

103    We address the second and fifth points further, and touch briefly on the sixth, in section 3.3 of these reasons.

3    SECTION 10(1) IN THE PRESENT CASE

104    In their respective Concise Statements and written submissions, the parties agreed that the relevant human right for the purposes of this case was the “right to social security” referred to in art 5(e)(iv) of the Convention. Beyond that, there was some divergence as to the correct approach. The applicant focused on the right to receive the age pension, which is (it was agreed) an aspect of the implementation of the human right to social security. The Commonwealth argued that, in order to determine whether Indigenous people enjoy the human right to social security to a lesser extent than members of other races, it was necessary to examine the operation of the social security system as a whole.

3.1     The right to social security

105    In the light of the way “right” is defined in s 10(1), and the statements in Maloney referred to at [24] above, we agree that the relevant human right in the present case should be understood to be “the right to public health, medical care, social security and social services” (art 5(e)(iv) of the Convention).

106    It is human rights, identified at the level of generality seen in art 5 (and distinguished from the specific rights enforceable under municipal law), to which the Convention is directed. The language of art 5 indicates that the human rights listed there are, so far as the Convention is concerned, examples of a broader genus (Maloney at [300] (Gageler J)). However, it is sufficient for present purposes to focus on the right to “social security”.

107    “Social security”, at least in the English speaking world, is usually understood to refer to a state-run system providing financial support to people who are unemployed, sick, retired, or otherwise in need (see, eg, the Oxford English Dictionary). The enjoyment of the “right” to social security will therefore usually depend on government provision, through legislation or otherwise. In Australia, “social security” is not a common law right or a fundamental freedom that is limited to a greater or lesser extent by legislation. To the extent that it is embodied in rights enforceable under Australian law, those rights are statutory. Equal enjoyment of the right to social security in Australia depends on the extent to which legislation confers rights under Australian law to income (and potentially other) support, not the extent to which legislation prohibits or penalises conduct.

108    The concept of “social security” is not defined in the Convention. Having opened for signature in December 1965, the Convention pre-dates the International Covenant on Economic, Social and Cultural Rights, opened for signature 16 December 1966, 993 UNTS 3 (entered into force 3 January 1976) (ICESCR) (which recognises “the right of everyone to social security” in art 9); so that, while their drafting processes no doubt overlapped and may have informed each other, the ICESCR is not available as a tool for construing the Convention in so far as the RDA adopts it as part of Australian law (and in any event “social security” is not defined in the ICESCR). The discussions of the concept of “social security” in the international law sphere to which we were referred relate to the ICESCR and postdate both the Convention and the RDA. They are therefore not available as “extrinsic materials” for the purpose of construing s 10 (Maloney at [15] (French CJ), [61] (Hayne J), [134] (Crennan J), [235] (Bell J)), although they “may usefully draw attention to possible arguments about how the RDA should be construed” (Hayne J at [61]). In any event, there does not appear to be any real disagreement as to what the concept involves. The United Nations Committee on Economic, Social and Cultural Rights observed in its General Comment No 19: The right to social security (art. 9), 39th sess, UN Doc E/C.12/GC/19 (4 February 2008) at [2]:

The right to social security encompasses the right to access and maintain benefits, whether in cash or in kind, without discrimination in order to secure protection, inter alia, from (a) lack of work-related income caused by sickness, disability, maternity, employment injury, unemployment, old age, or death of a family member; (b) unaffordable access to health care; (c) insufficient family support, particularly for children and adult dependents.

109    The same Committee, in its views on an individual case published in 2018, noted that the “right to old-age benefits” was recognised as part of the “right to social security” which art 9 of the ICESCR required to be satisfied at “at the very least, minimum essential levels” (Committee on Economic, Social and Cultural Rights, Views: Communication No 10/2015, UN Doc E/C.12/62/D/10/2015 (26 March 2018) at [11.2]–[11.3] (Trujillo Calero v Ecuador)). Again, this is not itself determinative as to the construction of art 5 of the Convention, but reflects what we understand to be an uncontroversial view as to what “the right to social security” entails.

110    There may be room for the exercise of judgment by individual States as to what should be covered by a regime of social security, as well as the level of provision. Such questions do not arise here, because the Convention does not require effectuation of the “right to social security” to any particular extent; rather, it requires states parties to guarantee that the right is equally enjoyed by members of all races. The point is that “social security” connotes a regime for the provision of benefits protecting people against a lack of work-related income arising from a range of causes.

111    Consideration for the purpose of s 10(1) of the extent to which people of different races in Australia enjoy the right to social security involves, primarily, analysis of how the SSA applies to them. It is the primary means by which the Commonwealth seeks to give effect to that human right. In principle, as the Commonwealth submitted, all of the pensions and benefits provided for by the SSA (and possibly other legislation as well) are relevant to an understanding of whether people of a particular race enjoy “the right to social security” to a lesser extent than people of another race. That is because the right is broadly expressed and is given effect under Australian law by the combination of all of those pensions and benefits. However, in our view, this proposition does not assist the Commonwealth in the present case, for reasons we will shortly attempt to explain.

112    Before doing that, it is necessary to give a brief account of the pensions and benefits for which the SSA provides.

113    The age pension, which is central to the issues in the present case, has the following features.

(a)    Qualification for the pension involves attaining a defined age (67 years for men and women in the applicant’s age group: SSA s 23(1)) and satisfying either a residence criterion or one of a set of alternative criteria (which are not presently relevant): SSA s 43.

(b)    The calculator used to determine the rate of pension that is payable per fortnight depends on whether the person is blind (s 1065) or not (s 1064). In either case the calculation is fairly complicated and involves several steps. The rate depends principally on the person’s income and assets (and, if applicable, the income and assets of their spouse or de facto spouse). The dollar amounts specified in ss 1064 and 1065 are adjusted six monthly by processes provided for in Pt 3.16 of the SSA. If the rate calculated using the relevant calculator is zero, no pension is payable (s 44).

(c)    Qualification for the pension continues until the person dies. The amount that is actually payable (if any) during that period will change from time to time depending on rate adjustments, the person’s income and assets and any changes in their relationship status.

114    The other pensions and benefits provided for by the SSA are, briefly, as follows.

(a)    Part 2.3 of the SSA provides for the disability support pension. It is intended to provide income support for people aged over 16 who have a continuing inability to work as a result of a physical, intellectual or psychiatric impairment. The rate of the disability pension is calculated using Pension Rate Calculator A in s 1064, which is one of the calculators potentially used to calculate the rate of the age pension. Thus, the recipient’s assets and ordinary income are taken into account in calculating the rate that is payable.

(b)    Part 2.5 of the SSA provides for the carer payment. It is intended to provide income support to a person who cannot support themselves through paid employment due to the demands of caring for another person. The rate of the carer payment is also calculated using Pension Rate Calculator A.

(c)    Part 2.10 of the SSA provides for the parenting payment. It is intended to provide income support for the parent of a child who is that child’s principal carer until the child reaches a specified age (six years if the recipient is in a couple, and eight years if they are not). The rate is calculated in accordance with the Pension PP (Single) Rate Calculator at the end of s 1068A or the Benefit PP (Partnered) Rate Calculator at the end of s 1068B as the case may be. Both of these calculators make the rate of income support payments dependent on the recipient’s assets and ordinary income, although they contain some incentives for recipients to participate in the paid workforce.

(d)    Part 2.11 of the SSA provides for youth allowance. It provides income support to people aged from 16 to 21 who are looking for employment, and to full-time students and apprentices aged 16 to 24. The rate of youth allowance is calculated in accordance with the Youth Allowance Rate Calculator in s 1067G. As with other benefits, this calculator makes the rate of the allowance dependent on ordinary income and assets.

(e)    Part 2.11A of the SSA provides for Austudy. Austudy provides income support for full-time students and apprentices who are older than the maximum age for youth allowance. The rate of Austudy for a person who is eligible is calculated using the Austudy Payment Rate Calculator in s 1067L, which makes the rate of payment dependent on the recipient’s ordinary income. An assets test is also applicable.

(f)    The jobseeker payment, which is provided for by Pt 2.12 of the SSA provides income support for people aged between 22 years and the pension age who are unemployed. The rate of job seeker payment is calculated using Benefit Rate Calculator B at the end of s 1068, which also takes into account the recipient’s ordinary income. An assets test is also applicable. To maintain eligibility for the jobseeker payment, a recipient of the payment is required to undertake “mutual obligation” requirements per s 593(1AC) of the SSA, by satisfying their “employment pathway plan” (EPP) to demonstrate their willingness to actively seek, accept and undertake suitable paid work. What the mutual obligation requirements entail varies depending on age, assessed work capacity, and parental responsibilities, but at a general level an EPP may contain requirements relating to the person actively seeking suitable paid work in Australia, accepting, or being willing to accept, offers of suitable paid work, and undertaking, or being willing to undertake, suitable paid work. It may also require a person to undertake training or study, voluntary work, employment programs, interviews with employment services providers, and work on resumes (see s 40G of the Social Security (Administration) Act 1999 (Cth) (Administration Act)).

(g)    Part 2.15 of the SSA provides for the special benefit, which may be made at the discretion of the Secretary to a person who does not qualify for any other income support benefit and who is for any reason “unable to earn a sufficient livelihood for the person and the person’s dependents”. The rate of special benefit is in the discretion of the Secretary, but is not to exceed the rate at which youth allowance, Austudy or the jobseeker payment would be paid to the person if they were eligible for one of those payments.

115    An important point that emerges from this summary is that the age pension has characteristics which are likely to make it more attractive than most of the other payments potentially available under the SSA. It is paid at a higher rate than the jobseeker payment, the parenting payment, and Austudy. Perhaps more significantly, upon attaining the relevant pension age, eligibility to receive the age pension is assured for the rest of a person’s life. There is no need for an age pensioner to establish that they have any particular disability, are responsible for caring for others, or are actively seeking employment. Inability to qualify for the age pension is therefore not fully compensated for by the right to apply for other forms of income support under the SSA, or even by actually being eligible for those forms of support. Accordingly, if the applicant succeeded in establishing that Indigenous men enjoyed the right to the age pension to a lesser extent than other men, that unequal enjoyment would not be cancelled out by access to those other forms of income support. The social security system as a whole would not treat members of all races with equal dignity and respect if it provided members of a particular race with more limited access to the age pension than others.

116    Some information is included in the special case at [94]–[95] indicating a high level of welfare dependency among Indigenous people. That material does not support any specific findings but does suggest that, if the issue for resolution was a broad one as to whether Indigenous people gain less from the social security system than people of other races, the applicant may not have met the onus of proof that he bears. As suggested from the bar table, it might well be found (given more complete evidence) that Indigenous men between 55 and 67 years of age are far more likely than other men to be receiving a pension or benefit under the SSA. It might well also be found that, having qualified for the age pension, Indigenous men on average receive the pension at a higher rate than other men.

117    However, those facts, if true, simply reflect the greater incidence of disadvantage and poor health (relative to other Australians) among Indigenous people and the application of the SSA to those circumstances. An approach to the question emphasising ‘swings and roundabouts’ is unattractive if the roundabouts constitute no more than the social security system responding to the poverty and poor health suffered by Indigenous people. The figures on welfare dependency are simply the result that is to be expected if the provisions of the SSA are racially neutral (as they are) and make eligibility and rates of payment dependent on need (as they do). They do not indicate any greater enjoyment of any aspect of “the right to social security” by Indigenous Australians compared to others. They do not provide an answer to the applicant’s argument concerning access to the age pension.

118    It is therefore necessary, as the applicant urged, to focus attention on the age pension itself in order to decide whether any lesser enjoyment of the human right to social security arises from the stipulation of a single “qualifying age” for all applicants. The issue can be framed in terms of the “right” to receive the age pension, although it needs to be kept in mind that the relevant “right” for s 10 purposes is the human right to social security.

3.2     The right to receive the age pension

119    The corollary of focusing on the right to receive the age pension, rather than bringing to account the extent to which members of different races in fact receive the various benefits provided under the SSA, is that the content and characterisation of that right become central to the assessment of whether, in view of their shorter life expectancy, Indigenous men, taken as a group, enjoy the relevant human right to a lesser extent than men who belong to other races. The content of the right to the age pension is, of course, found in the statute. The essential features of the age pension have been noted above.

120    A pension of substantially this nature has been part of Australia’s social welfare system since the enactment of the Invalid and Old-Age Pensions Act 1908 (Cth) (1908 Act). Old-Age Pension Acts had been enacted by New South Wales in 1900 and by Victoria in 1901. The Attorney-General, the Hon. Littleton Groom MP, described the reasons for introduction of a Commonwealth Age Pension in this way, in his second reading speech to the Bill for the 1908 Act (Commonwealth, Parliamentary Debates, House of Representatives, 3 June 1908, 11922):

In every enlightened community the establishment of old-age pensions is regarded as an ideal whose attainment should be earnestly sought, it being felt to be a reproach to civilisation that many persons whose lives have been spent in working for the advancement of the State should in their old age, through no fault of their own, be compelled to end their days in charitable institutions.

121    The 1908 Act was amended numerous times before being replaced by parts of the Social Services Consolidation Act 1947 (Cth) (later renamed the Social Security Act) (1947 Act), which brought together provisions for a number of pensions and benefits (including unemployment benefits, invalid pensions, widows’ pensions and benefits for single parents). The 1947 Act was also amended numerous times before being replaced by the present SSA. Through that long history of amendments there has been provision for a pension, referred to as the “old-age” or “age pension”, for which the basic qualification has been the attainment of a specified age in a person’s 60s which, in general terms, corresponds with an age at which most people would be expected to be no longer in the paid work force.

122    While qualification by way of age has been a constant feature, other aspects of the pension have changed. The required period of residence in Australia has changed. Different approaches have been taken at different times to whether there should be a different rate for married pensioners and how other income and assets should be accounted for.

123    Initially, the qualification for the pension expressly excluded “aliens”, “asiatics (except those born in Australia)” and “aboriginal natives of Australia, Africa, the Islands of the Pacific or New Zealand” (1908 Act, s 16). These exclusions were gradually removed from 1926. Indigenous people became eligible for the old-age pension under s 13 of the Invalid and Old-Age Pensions Act 1942 (Cth) if they were exempted from State laws relating to the “control of Aboriginal natives” or, in States where such an exemption was not available, depending on an assessment of their “character, standard of intelligence and development”. Qualification was extended under s 24 of the Social Services Act 1959 (Cth) to all Indigenous people who did not follow a “nomadic or primitive life”; and the exclusion of Indigenous people was finally ended by s 29 of the Social Services Act 1966 (Cth).

124    The qualifying age for the old-age or age pension has remained relatively stable over the course of 115 years. Under s 15(1) of the 1908 Act, every person over 65 years of age was qualified along with everyone who was over 60 and permanently incapacitated; and the Governor-General was empowered by s 15(2) to alter the qualifying age for women to 60 years. The 1947 Act fixed the qualifying age for men at 65 years and for women at 60 (s 21). Section 21 of the SSA as originally enacted defined “pension age” as 60 years for women and 65 years for men. The two significant changes to the qualifying age have been as follows.

(a)    The Social Security Legislation Amendment Act (No 2) 1994 (Cth) (1994 Act) amended s 23 of the 1991 Act so that the “pension age” for women would gradually increase from 60 (for a woman born before 1 July 1935) to 65 years (for a woman born on or after 1 January 1949), thus aligning the pension ages for men and women from 1 January 2014.

(b)    The Social Security and Other Legislation Amendment (Pension Reform and Other Budget Measures) Act 2009 (Cth) (2009 Act) made further reforms to the definition of “pension age” in s 23. The pension age for men was to increase gradually from 65 (for men born up to 30 June 1952) to 67 years (for men born on or after 1 January 1957). The pension age for women was further amended so that it would continue to rise and would be 67 years for women born on or after 1 January 1957.

125    We were taken by counsel to parts of the extrinsic materials for the 1994 and 2009 Acts. We found this to be of some, but limited, assistance. The exercise upon which we are engaged in this part of the reasons is not the construction of the SSA, and does not focus on a search for legislative intention in the way that an argument about statutory construction would. The legal effect of relevant sections of the SSA is tolerably clear without the need for recourse to extrinsic materials. Rather, the task in this part of the reasons is to seek to characterise the right provided by those sections that confer the entitlement to the age pension, in order to come to a view about whether the stipulation of the same qualification age for all persons who otherwise qualify should be regarded as causing Indigenous people to “enjoy” the human right to social security to a lesser extent than others.

126    If being qualified to receive the age pension for a greater or lesser period of time is regarded as constituting greater or lesser enjoyment of the right to receive the pension, the applicant’s submissions clearly have considerable force. However, we do not think the right is properly characterised in that way.

127    The age pension has never been understood to provide support for a term of years or in a particular amount of money. Rather, as embodied in the legislation, the age pension and its predecessor the old-age pension, have always provided protection against falling into poverty in retirement. It is inherent in its design that the period for which a person qualifies for the pension will depend on whether they reach the relevant age and how long they live afterwards. The pension reflects an understanding that old age is a period during which a person is not only likely to be unable to earn a living by working, but is not obliged to try to do so; so that an income sufficient to provide a minimal level of dignity in retirement is provided for by the state. That understanding is reflected in the fact that qualification for the old-age and age pensions, for persons above the relevant pension age, has never depended on establishing that an individual applicant is unfit to work or unable to find work. It is also reflected in the extract from the second reading speech for the 1908 Act set out above.

128    The adjustments of the “pension age” enacted by the 2009 Act grew out of concerns about the “sustainability” of the age pension — that is, its cost to the budget — in circumstances where demographic change in Australia was expected to mean that the ratio of working age people to age pensioners would decline substantially over the next 40 years. It was pointed out that in 1909, when the old-age pension came into effect, around half the male population reached “retirement age” (ie 65). That figure had risen to 85 percent by 2009, and the life expectancy of men at retirement age in 2009 was seven years longer than it had been in 1909: see Budget Paper No 1 (2009-10) at 1–36 and the second reading speech to the Bill for the 2009 Act (Commonwealth, Parliamentary Debates, House of Representatives, 15 June 2009, 5853 (Ms Macklin)). The response, reflected in the amendments made by the 2009 Act, was to raise the “retirement age”; ie the age from which people with no or limited income from other sources could expect the Commonwealth to supplement their income for the rest of their lives, without any need to prove invalidity or to be available for paid work. This was consistent with the overall character of the age pension referred to in the previous paragraph.

129    The references to “retirement age” in the materials discussed in the previous paragraph indicate that the age pension may itself have influenced expectations and understandings in Australia as to the age at which people will retire from paid work. The Women’s Budget Statement for the 1993–94 financial year suggested that women were disadvantaged by having to fund a longer period of retirement than men, as a result of expectations that they would retire at 60 years of age: Commonwealth, Women’s Budget Statement 1993–4 at 230. A similar point was noted by Dr Jeff Harmer in a review which formed part of the basis for the changes made in the 2009 Act: Jeff Harmer, Pension Review Report (27 February 2009) at xix.

130    The “right” conferred by the provisions in the SSA that provide for the age pension should therefore be understood as a right to a level of income support, covering the period from when a person reaches “retirement age” until death, however long that period might be. The period (if any) for which that support is available to an individual depends on the individual’s lifespan. A person who lives for a relatively short period after retirement age (eg a person who dies unexpectedly), and thus needs (and qualifies for) that support for a relatively short period, does not enjoy that statutory right to a lesser extent than somebody who lives for a longer period. It is equally available to each of them, for the duration of their respective lives after they qualify. The relevant connection is with the duration of their lives after becoming entitled to receive the pension, not with their race.

131    Thus, an Aboriginal man in Australia does not enjoy the “right” to an age pension on a narrower basis or to a lesser extent than a non-Aboriginal man of the same age. Rather, all that can be said is that, based on the statistics, a randomly selected Aboriginal man will probably qualify for the pension for a shorter period than a randomly selected non-Aboriginal man of the same age. Each enjoys the right to the pension (subject to meeting payability criteria) for as long as he lives and thus as long as he needs it.

132    This situation is to be contrasted with, for example, provision for government-funded medical services excluding a serious condition associated only or predominantly with members of a particular race, such as rheumatic heart disease. In that hypothetical example the criterion of availability is facially neutral but there is a connection between a gap in the comprehensiveness of the care regime on the one hand and race on the other. The inequality in the present case lies in the fact that Indigenous men as a group spend statistically less time in old age than men of other races, rather than in the way the age pension provisions operate upon those differing lifespans. The difference in life expectancy is the consequence of a host of factors beyond the contemplation of the SSA; and each of those factors, as noted above, reflects social disadvantage rather than something intrinsic to Indigenous people. Those aspects of disadvantage and the resulting difference in life expectancy may change over time, leading to different levels of access to the age pension while the content of the domestic law “right” remains constant.

133    This is therefore not a situation where (to return to Gerhardy) the impugned law “omits to make enjoyment of the right universal, ie by failing to confer it on persons of a particular race” (or conferring it unequally). When the age pension is properly understood, it can be seen that the provisions conferring it do not result in a lesser enjoyment of that domestic law right — or, correspondingly, the relevant human right — by members of any race compared to other persons.

134    Therefore, the applicant has not demonstrated any lesser enjoyment of the relevant human right — the right to social security. It follows that question 2 in the special case must be answered “no”.

3.3     The framing of appropriate relief and the scope of s 10

135    If (contrary to our reasoning above) question 2 in the special case were answered “yes”, the outcomes suggested in question 3 at least broadly indicate the effect that s 10(1) would be taken to have. Section 10(1) would operate so as to augment the conferral of the relevant domestic law right (presumed in question 2 to be “the right to apply for and receive the age pension”) on Aboriginal and Torres Strait Islander men to such an extent that they would no longer enjoy that right “to a more limited extent” than non-Indigenous men.

136    If (as the applicant submitted) the reason for that “more limited” enjoyment is that Indigenous men (on average) enjoy the benefit of the age pension for a shorter period, the effect of s 10(1) must be to extend the period for which they are entitled to apply for and receive the pension. That can only be done by substituting for the statutory “pension age” another age that cancels out the disparity. On the agreed facts before the Court that would, at least at present, be 64 years. This reasoning necessarily applies to all Indigenous men in the relevant age group. Whether the question was formally answered in terms referring to “the Applicant” or “the represented persons” would depend on the issue to be considered later as to the representative status of the proceeding.

137    We noted above that reliance on evidence or agreed facts in the form of statistics or probabilities, drawn from research in the social sciences, is not necessarily fatal to the success of an argument relying on s 10(1), but that attention needs to be paid to what that evidence or those facts actually establish.

138    Here, the agreed facts as to the life expectancies of Indigenous and non-Indigenous men establish statistically-based expectations as to how long men in various age cohorts will live. Self-evidently, they are predictive and therefore involve guesswork. Also self-evidently, they do not take account of individual variations in physiology, lifestyle and risk factors. They could not form a basis for a finding that any individual is going to die at a particular time. The usefulness of facts of this nature depends, in part, on the propositions ultimately sought to be established.

139    At the stage of considering the effect of s 10(1), in a case such as the present, the utility of agreed facts based on life expectancy reaches its limit. That is because, unlike a provision such as art 15 of the Canadian Charter of Rights and Freedoms, which invalidates laws that contravene a principle of equality, s 10(1) in the operation with which we are presently concerned operates to remediate an identified gap in the conferral of rights.

140    Life expectancy figures, as noted above, tell us nothing about the circumstances or lifespan of any individual, nor of the individuals who make up the group as defined. The figures reflect the interplay of a host of factors affecting the lives of large numbers of people. They are an estimation of the future effects of those factors on members of particular groups, rather than something characteristic of membership of a race. The statistics tell us that, in the cohort of Indigenous men to which the applicant belongs, there are some whose lives will be shortened by three years or more as a result of factors arising from their race; there are others whose lives will be shortened by less than three years by those factors; and there are likely others whose lives will not be shortened in any measurable way. The operation of s 10(1) proposed by the applicant — treating all Indigenous men in his age group as reaching pension age when they turn 64 (or some other age identified by the Court) instead of 67 — does not adjust the SSA’s conferral of the age pension in an adequate or principled way.

141    On one view, what we have just said is a point going to the framing of relief (which is not presently before us) or, at most, to the correct answer to question 3 in the special case. However, in our view, the problem runs deeper. Because nobody knows when they will die, if the applicant’s construction of s 10(1) is correct, framing an appropriate order in the present case — even at an individual level — is impossible. The impossibility of identifying what s 10(1) would actually do, if the application of one pension age to all Australians is regarded as causing Indigenous people to “enjoy” a human right “to a lesser extent” than other people, suggests that the provision should not be construed in that way. The notion of unequal enjoyment of a human right should not be regarded as engaged by a law granting a pension or other benefit by reference to racially neutral criteria, if its asserted flaw is no more than the failure to be designed in a way that compensates some individuals accurately for the longer term adverse socio-economic effects of race.

142    This is a further reason why, in our view, the applicant’s submissions should be rejected and question 2 answered accordingly. The fact that Indigenous men in the applicant’s age group are expected on average to live — and thus enjoy access to the age pension — for around three years less than other men is, as the agreed facts make clear, a tragic consequence of two centuries of dispossession, marginalisation and destruction of social structures. It does not follow from that fact that the SSA, in its practical application, is inconsistent with “equality before the law” in a way that engages s 10(1). The unacceptable differences in health and socio-economic status between the people of Australia’s first nations and others is clearly a matter that is appropriate to be addressed by measures that would come within s 8(1) of the RDA. However, the inclusion of s 8 in the RDA (reflecting art 2.2 of the Convention) indicates that Parliament intended to maintain for itself and the State and Territory legislatures scope to decide the extent and nature of such measures. It weighs against construing s 10(1) as a general provision that, by its own force, modifies existing legal regimes in an attempt to equalise outcomes.

143    The problem of identifying the form of declaratory or other relief that properly reflects the effect of s 10(1) in a case such as the present one may possibly point to a broader objection to the construction propounded by the applicant. In a situation such as the one envisaged in Gerhardy at 98 (where the legislature “omits to make enjoyment of the right universal, ie by failing to confer it on persons of a particular race”), it will probably be fairly clear that the relevant law is to be administered as if the right had been conferred universally. However, if the domestic law right is not withheld on some express basis that is clearly linked to race, identification of the additional rights that s 10(1) confers in order to bring about the equal enjoyment of the relevant human right may involve complex and contestable judgments, based on potentially shifting facts. Such judgments could test the limits of what a court exercising the judicial power of the Commonwealth may properly do. Clearly, how that potential problem might bear upon the construction of s 10(1) is itself a complicated issue. In the light of the conclusions we have reached above, it is unnecessary to express a view on that issue.

3.4     A final point

144    The applicant’s case, reflected in the declarations that he seeks, is that he and the represented persons have a right to receive the age pension that arises pursuant to s 10(1). That reflects an orthodox approach in the light of the High Court authorities referred to above which hold that, rather than modifying or overriding laws that confer rights, s 10(1) supplements them by conferring additional rights.

145    During the hearing the Court raised with counsel for the applicant a question as to whether there is an appropriation that would allow amounts to be paid to the applicant and the represented persons by way of pension if his submissions were accepted. By reason of s 83 of the Constitution, the availability of funds under an existing appropriation is a prerequisite to the payment of any money out of the Treasury of the Commonwealth even where there is otherwise an obligation to make such a payment.

146    Section 242 of the Administration Act appropriates the Consolidated Revenue Fund to the extent necessary to make “payments under the social security law”. The “social security law”, by reference to s 3 of the Administration Act and s 23(17) of the SSA, comprises these two Acts (and legislative instruments made under them) and any other Act or provision “expressed to form part of the social security law”. It does not, at least at first blush, include the RDA. It would no doubt be possible to read the “social security law” as including the SSA as construed with or modified by the RDA, but that is not how s 10(1) of the RDA has hitherto been described as operating in a case where the impugned law is one that confers rights.

147    It is not appropriate to express a concluded view on this point. It is not necessary in order to answer the questions posed by the special case and it has only been addressed very briefly by the parties. If our answers to questions 2 and 3 were different, the point would need to be addressed in the course of considering what, if any, injunctive relief should be granted.

4    REPRESENTATIVE PROCEEDING

148    As noted earlier, the proceeding is currently framed as a representative proceeding in reliance on Pt IVA of the FCA Act and r 9.21 of the Rules. The represented group is described in the Further Amended Originating Application as:

Indigenous men who:

(a)    will turn 65 years of age in 2022;

(b)    are men whose “pension age”, for the purposes of the [SSA], is determined by item 5 of the table in s 23(5A) of that Act and who will therefore satisfy the pension age requirement when they turn 67;

(c)    satisfy the other qualification criteria in s 43 and the payability criteria in ss 44-47A of the [SSA];

(d)    in the event that the Applicant succeeds in establishing in this proceeding that, by reason of the application of s 10 of the [RDA], the pension age for the Applicant and the represented persons was not 67 years of age as specified in s 23(5A) of the [SSA], but, rather, was:

(i)    64 years of age; or

(ii)    some other age less than 67 years of age, they [sic] will:

(iii)    satisfy the pension age requirement upon turning 64 years of age or some other age less than 67 years of age; and

(iv)    will be entitled to be paid the age pension if they make a claim for it and at the date of the claim they continue to satisfy the other qualification criteria in s 43 and the payability criteria in ss 44-47A of the [SSA].

149    Rule 9.21 provides that a proceeding may be started or continued by a person “as representing all or some of the persons who have the same interest and could have been parties to the proceeding”. The Commonwealth submits that the group of “represented persons” identified by the applicant does not meet that requirement.

150    Question 1 in the special case asks, in effect, whether that submission is correct. (The question does not ask this Court to determine whether as a matter of discretion the proceeding should continue as a representative proceeding (cf O’Donnell v Commonwealth [2021] FCA 1223 at [18]). That is clearly a matter for consideration by the docket judge.)

151    It is said that the requirement for a common interest in the proceeding is not met for the reason identified in Campbells Cash and Carry Pty Ltd v Fostif Pty Ltd [2006] HCA 41; 229 CLR 386 (Campbells Cash and Carry); that is, that membership of the group depends on the occurrence of a future event. In Campbells Cash and Carry the group was defined to include only members of a larger class who “opted in” to the proceeding. No member of the larger class had any interest in the proceeding (other than the precedential value of a judgment) unless and until they took that step. In effect, therefore, the group necessarily had no members at the time of commencement of the proceeding.

152    The group definition in the present case does not have that vice. Membership does not (as the Commonwealth appears to submit) depend on the making of a claim for the age pension (a future event). Rather, one of the characteristics of a group member is that he is someone who, if the applicant’s submissions as to the effect of s 10(1) are accepted, will be entitled to be paid the age pension if certain conditions are satisfied. That expresses an existing characteristic rather than conditioning group membership on something happening in the future.

153    Paragraph (d)(iv) of the definition of the group superficially suggests complexity because it was impossible to know when the proceedings were commenced whether an individual who came within paras (a)–(c) and (d)(iii) would in due course make a claim and whether on making that claim he would satisfy the “other qualification criteria” and the “payability criteria”. However, we do not consider that to be fatal to the representative nature of the proceeding. The group identified is, in substance, men who:

(a)    were born in 1957;

(b)    therefore, according to the terms of the SSA, will not reach “pension age” (and therefore cannot meet the qualification criteria for the age pension) until they reach 67 years of age (which will happen in 2024); but

(c)    if the applicant’s submissions concerning s 10(1) are accepted, reached pension age when they turned 64 (which occurred in 2021); and

(d)    when the proceedings were commenced, satisfied the other qualification criteria and the payability criteria.

154    Members of that group have an interest in the proceeding which, while its practical value is contingent on other things (ie making a claim and meeting the other qualification criteria and the payability criteria), is in principle the same as that of the applicant. If his argument is upheld, they meet one of the qualification criteria for the age pension; if it is not, they will not meet that criterion until 2024. The pension age is identified in the SSA as an objective criterion, not a matter for assessment by a decision-maker, so that meeting or failing to meet that criterion has a legal consequence. It is not the case (as submitted orally) that the applicant can only represent persons who have made claims and had them rejected.

155    It may therefore be that the last criterion set out in our summary above (which reflects para (c) of the applicant’s definition) narrows the class unnecessarily. It also appears us that para (d)(iv) of the applicant’s definition does no more than state an obvious consequence of what precedes it. However, these observations do not mean that the members of the class do not have the requisite common interest.

156    We would therefore answer question 1 “yes”.

5    ANSWERS TO QUESTIONS

157    We therefore answer the questions of law in the special case as follows:

(1)    Yes.

(2)    No.

(3)    Does not arise.

158    The special case does not ask this Court to decide any question as to costs. Questions concerning the costs of the special case and the further conduct of the proceeding are matters upon which the parties should be given an opportunity to address, and if possible, agree orders. That may need to await the expiry of the time period for any application for special leave to appeal.

Associate:

Dated:     12 July 2023

ANNEXURE