Sydney Law Review
There are many words which have been made to suffer constant misuse; but there is none which suffers more abundantly, or with sadder consequences, that the word Race.
Ernst Barker 
The meaning and scope of the race power – s51(xxvi) – is unsettled. In the sole case in which the High Court was required to directly consider the issue, Kartinyeri v The Commonwealth  (the ‘Hindmarsh Island Bridge Case’), it was unable to reach a majority view on the provision’s meaning, and the tests that should be applied for deciding statutory validity under the provision.  This article considers the meaning and scope of the provision in the context of the decision. It argues that a critical term for understanding s51(xxvi)’s meaning is ‘race’, which appears in the provision.
This article also argues that the literalist method for constitutional interpretation is inadequate to the task of interpreting s51(xxvi) because of the complex nature of the term ‘race’. Literalists may too readily assume that the term has an immediately discernible plain and natural meaning. But the term’s apparent plainness is deceptive. Its complexity and truer meaning can only be appreciated in its social and historical context. For that reason, an inquiry into the meaning of the race power requires considering the historical circumstances of the creation of the provision and the meaning that was attached to race. The term ‘race’ then needs to be compared with the historical context of the provision’s amendment in 1967, and with the meaning attributed to the term today.
Once the meaning of ‘race’ is considered, the meaning of s51(xxvi) itself can be better understood. It is then necessary to devise a test for determining the validity of statutes relying on s51(xxvi) as a source of power. The test needs to maintain fidelity to the express words of s51(xxvi) while giving effect to its underlying meaning. A number of the members of the Court in the Hindmarsh Island Bridge Case proposed tests for determining constitutional validity. It is submitted that the one that best achieves the competing objectives of textual fidelity and giving effect to underlying meaning is that proposed by Gaudron J. Her test has an antecedence in the United States and Canadian equal protection jurisprudence. Although s51(xxvi) is not an equal protection clause, it parallels the equal protection requirements by requiring (it is argued) a rational connection between the process of deeming a group of people a race and the legislative measures regarding the group. In other words, the characteristics that purport to define a group as a race must be the rational subject of legislative attention. This proposition is more fully canvassed in Part 6.
Part 2 considers the various methods the High Court has adopted in interpreting the Constitution and argues that a literalist approach is unsuited to interpreting s51(xxvi). It is argued instead, in Part 3, that the broader historical circumstances of the creation and evolution of the provision must be considered. Part 4 proceeds to consider the historical circumstances of the creation of the race power, and the general meaning that was attached to the term ‘race’ at Federation. Part 5 compares the meaning of the term at Federation with that given in 1967 when the race power was amended, and the current meaning given to ‘race’. Finally, Part 6 considers the appropriate test for applying s51(xxvi).
A discussion of the interpretation of s51(xxvi) requires some common understanding of the appropriate method of interpreting constitutional provisions. The High Court has employed a number of methods for Constitutional interpretation.  Williams and Bradsen argue that the Court has adopted three methods for interpretation: original intent, textualism and ‘living force’, and that the Court has not settled on one method to the exclusion of the others.  There are elaborations on these basic methods, for example the often misused connotation/ denotation test, which is a variant on textualism. Under the test a term in a provision may connote a particular meaning, which theoretically remains constant from the time the provision commenced, but denotes things that were not contemplated at commencement. For example, in 1901 the term ‘jury’ under s80 of the Constitution denoted ‘a panel of men convened to decide questions of fact’. Today it denotes ‘a panel representing the community to decide questions of fact’.  At the time the Constitution was written juries did not denote a panel including women, but today they do. 
In the broadest sense the object of interpretation is ‘to give effect to the intention of the [law-maker] as that intention is to be gathered from the language employed having regard to the context in connection with which it is employed’  The debate about the most appropriate method for achieving the object has to a large extent turned on matters of emphasis. So-called literalists (or textualists) search for author intent, but emphasise that this is revealed in the textual expression chosen by the authors. So-called originalists place more emphasis on determining intention by examining the text within the context of the historical and other circumstances of the writing of the text. Those who subscribe to the ‘living force’ approach emphasise that the Constitution was written for the purpose of being relevant to many future generations and therefore the text is to be considered in the light of its current relevance and applicability. But the purposes underlying the creation of these categories are not mutually exclusive of each other. A literalist concedes that text cannot be understood devoid of its context; an originalist concedes that the text is highly significant in determining original intent; and those who emphasise that the Constitution is a living document concede that the task of interpreting it in a relevant and meaningful way does not provide license to ignore or contradict the text.
The difference between the various interpretation methods is not limited to issues of emphasis, they also differ on the bounds of relevant evidence in searching for intent. A literalist will generally permit less evidence for determining author intent than an originalist. A literalist will be reluctant to look beyond the written text of the Constitution and judgments that have considered the written text. An originalist is generally keen to look beyond the text to historical and other circumstances surrounding the writing of the text. These boundaries are not as fixed as may first appear, and will often vary with the circumstances. For example, procedural provisions like the one dealing with the retirement age for High Court judges  , are less likely to involve the need to proceed particularly far from the bounds of the text regardless of which interpretation method is applied. Broader and more obscure provisions, like the race power, are likely to require relying on more than the text for discovering the underlying intention of the text. This may explain to some extent why the High Court has used the various interpretation methods interchangeably without rigorously sticking to any one method.
How then should s51(xxvi) be interpreted? In the Hindmarsh Island Bridge Case, members of the Court, unsurprisingly, differ on the appropriate interpretation method. Gummow and Hayne JJ tend towards a literalist approach while Gaudron J and Kirby J each rely less on literalism. Gaudron J reads s51(xxvi) in broad contextual terms, while Kirby J is prepared to consider the historical context in which the provision was created and amended.  Brennan CJ and McHugh J avoid directly confronting the issue of constitutional interpretation as they consider it is not at issue. Gummow and Hayne JJ, in inclining towards a literalist approach, emphasise that the constitutional text must always be controlling.  Kirby J also emphasises the significance of the constitutional text by stating that the ‘text is the law’  , but does not see that paying due homage to the text should detract from examining the historical circumstances of the creation of the text.  Gummow and Hayne JJ underline their literalist tendencies by expressing doubts about the necessity for considering the circumstances of the enactment of the 1967 referendum legislation for determining the meaning of the amended race power. 
Literalists are reluctant to proceed beyond the bounds of the text because they consider that ‘the Constitution is to be interpreted by reading its words according to their natural sense and in documentary context, and then giving to them their full effect’.  Literalists assume that author intent can generally be revealed in the plain words the author chooses to express his or her intent. One problem with this is that it assumes that the author and interpreter will broadly share the same world view. That is, in a general sense, that a shared view exists between the author and the interpreter as to what the relevant words mean. This assumption, however, becomes more problematic with time. The more distant the time between writing and interpretation, the greater the risk of a changed world view between author and interpreter leading to increased chances for ‘misinterpretation’. That is, the founders’ assumptions about the meaning of their written words in the Constitution may well not be shared by an interpreter a century or so later. It will be argued in this article that the race power is particularly prone to different assumed meanings by the authors and interpreters because the term ‘race’ has transformed in meaning, partly as a result of the traumatic experiences of the 20th century regarding race, and advancements in the pure and social sciences.
The problems arising from applying a literalist approach to s51(xxvi) can be demonstrated by searching for the plain and natural meaning of ‘race’. The term derives from the old French ‘rasse’ and Italian ‘razza’, which together with the Spanish and Portuguese ‘raza’ probably derive from the Arabic ‘rá’s’ meaning head, beginning, origin.  According to the Oxford English Dictionary, the term has been in regular use at least since the 16th century, and has been used to connote an extraordinary range of meanings, including ‘A group of persons, animals, or plants connected by common descent or origin’, ‘A limited group of persons descended from a common ancestor; a house, family, kindred’, ‘A tribe, nation, or people, regarded as common stock’, ‘A group of several tribes or peoples, regarded as forming a distinct ethnical stock’, and ‘One of the great divisions of mankind, having certain physical peculiarities in common’.  A final definition offered by the Dictionary is particularly wide, as it encompasses ‘A group or class of persons, animals, or things, having some common feature or features’. A subset meaning attributed to this definition is ‘A set or class of persons’ for which the Dictionary attributes usage of the term to a number of authors over the centuries who at various times have referred to a race of idle people, the race of learned men, the race of poets and the two races of men: the men who borrow and the men who lend. 
The extraordinary diversity of the term ‘race’ in Australian usage is confirmed by the Macquarie Dictionary which offers the following definition:
1. a group of persons connected by common descent.
2. a population so connected.
3. Ethnology a subdivision of stock, characterised by a more or less unique combination of physical traits which are transmitted in descent.
4. the state of belonging to a certain ethnic stock.
5. Zoology a variety; a subspecies.
6. a natural kind of living creature: the human race; the race of fishes .
7. any group, class, or kind, especially of persons .... 
It is, of course, not uncommon for a term to have multiple meanings, and usually common sense and experience will direct the interpreter to the most appropriate meaning for commencing the task of interpretation. The problem with ‘race’ is that a starting point is not as apparent as may first appear. Some hints at the difficulty of the task can be found in summary form in the Fontana Dictionary of Modern Thought under its consideration of the meaning of ‘race’:
race. A classificatory term, broadly equivalent to subspecies. Applied most frequently to human beings, it indicates a group characterized by closeness of common descent and usually also by some shared physical distinctiveness such as colour of skin. Biologically, the CONCEPT has only limited value. Most scientists today recognize that all humans derive from a common stock and that groups within the SPECIES have migrated and intermarried constantly. Human populations therefore constitute a GENETIC continuum where racial distinctions are relative, not absolute .... It is also acknowledged that visible characteristics, popularly regarded as major racial pointers, are not inherited in any simple package and that they reflect only a small proportion of an individual’s genetic make-up .... Socially, race has a significance dependent not upon science but upon belief. Men depict themselves and see one another in terms of groups which, however frail their objective basis, thereby assume social importance. 
These observations point to the lack of an objective reference point for defining race. That is, the term is socially determined, and is relative to the time and place where the determination is made. The highly relative (and inflammatory) nature of racial categorisations has been illustrated in recent times. The racial genocide conducted in Rwanda and Burundi involved the Hutu and Tutsi enthnic groups which were barely distinguishable, partly because intermarriage between the groups had been common. Both groups spoke the same language, could not be readily distinguished by appearance (although the Tutsi were usually taller) and both lived together in settlements before the genocide took place.  Banton notes that often the distinction between ‘racial’ or ‘ethnic’ groups can be barely transparent:
It was reported that when travellers were stopped at road blocks in Yugoslavia at the beginning of the conflict they might be asked to recite the Lord’s Prayer, because although they spoke a common language Serbs and Croats tended to use different words for “bread” in the plea “give us this day our daily bread”. In Rwanda people were required to carry an identity card which specified their ethnic group. 
The Nazis had similar problems identifying Jews. For although some Nazis claimed that Jews had different head and facial structures to non-Jews which confirmed their biological inferiority, the Nazis nevertheless required Jews to wear Star of David armbands and to carry identity papers so they could be distinguished.
This is not to suggest that different groups in society have distinct traditions, language, shared history and identity, but it does mean that the term ‘race’ lacks the objective and plain and natural meaning that may first be supposed. It is, in fact, a term which is highly relative in time and place. A group that may be thought of as a distinct racial group a century ago may not be considered as such today. Similarly, a group that may be thought of as a distinct racial group in one country may not be thought of as such in another. Thus there is a danger in a literalist assuming that their understanding of the plain and natural meaning of the term ‘race’ was shared by the Constitution’s founders. Literalism therefore faces the difficulty having to select the most appropriate of the multiple meanings of race. And it must offer some rational basis for its selection. Admittedly the task is not confined to literalists, but they undertake the task in more confined (and inadequate) contextual surroundings than non-literalists, which can lead them to error. Literalism may lead an interpreter to connote a meaning to the race which bears little relationship with its meaning when s51(xxvi) commenced. Avoiding this danger involves developing an understanding of the historical circumstances of s51(xxvi)’s creation.
If it is accepted that strict literalism will not suffice in the interpretation of s51(xxvi), the question arises as to how far one should go in examining extraneous material to gain insights into author intent to assist with interpreting the provision. It is possible, post Cole v Whitfield,  to examine the Constitutional Debates. However, as will be shown below, they alone offer scant insight into the minds of the founders. Griffith made a number of comments about the proposed race power which can only be adequately comprehended within the historical context in which they were made. The question then arises as to what extent we can proceed beyond the Debates to consider the historical context of the race power’s creation? Historical accounts are often contested, and lawyers will usually be reluctant to entertain alleged facts that cannot be reliably verified. There is also the risk of inviting a wealth of information which a court, in a practical sense, cannot properly digest. But these concerns must be weighed up against the risk of maintaining a studied ignorance of history, which could well lead to highly inaccurate readings of the Constitution’s text. As Goldsworthy observes: But once it is understood that the clarification of a statute’s meaning requires taking account of all relevant evidence of legislative intention, then it should be appreciated that there can be a wide variety of evidence, that some pieces of evidence may contradict others, and that a final judgment requires weighing them against one another. The difficulty of the task should not impugn its authenticity. 
In any event, the High Court has not avoided paying heed to history. In Cole v Whitfield  the Court felt compelled to return to basics to discover the original purpose of s92 to clear up the muddled and contradictory interpretations the Court had given the provision over preceding decades. The Court states that historical material may be referred to:
for the purpose of identifying the contemporary meaning of language used, the subject to which that language was directed and the nature and objectives of the movement towards Federation from which the compact of the Constitution finally emerged. 
History has been referred to in subsequent cases. In McGinty’s Case McHugh J states that the Constitution must be interpreted ‘according to the ordinary and natural meaning of its text, read in the light of its history’.  Reference to the historical sources is made in Victoria v Commonwealth where the history of treaty ratification and implementation in Australia towards the end of the 19th century is examined to interpret the scope of the external affairs power.  In Theophanous v Herald & Weekly Times Ltd McHugh J states that ‘each generation must read the provisions of the Constitution in their context and that includes the historical context of the Constitution’.  And in Cheatle v R the Court reviews historical evidence of criminal trials to decide the meaning of s80. 
In summary, the Constitutional Debates alone do not necessarily provide a sufficient historical context for ascertaining the meaning of constitutional provisions, and, it will be argued in the next section, do not provide a sufficient basis for understanding s51(xxvi). A deeper historical context is required to gain an adequate appreciation of the meaning attached to the term race at the time the provision was created, and to understand the underlying purpose of the provision.
The genesis of the race power was a proposal put by Sir Samuel Griffith to the 1891 Constitutional Convention to include an exclusive Commonwealth race power in the Constitution. His immediate purpose seems to have been for a power to deal with groups like the Kanakas in north Queensland.  Griffith had been involved with attempts to stop the ‘blackbirding’ of Pacific Islanders,  who were induced to enter a life of semi-slavery as labourers in the cane fields of Queensland, and were required to return home after three years.  Some Queenslanders were opposed to the use of Pacific Island workers because it would lead to moral contamination, while others were concerned that it involved elements of slavery.  The Queensland government attempted, however, to appease all sides with legislation, beginning with the Polynesian Labourers Act 1868, that provided for their segregation and for basic labour protection. 
The legislation, however, did not manage to prevent the Kanakas from being tricked into working for minimal sums on supposedly freely bargained contracts.  Griffith sponsored the Pacific Island Labourers Amendment Act 1885, which aimed at ending the use of the Pacific Islanders as indentured workers after 1890. The legislation coincided with the declining economic need for the Kanakas because of new technology and the recession of the 1890s during which there was a shortage of work. The practice was brought to a decisive end, however, when the federal Parliament enacted the Pacific Island Labourers Act 1901, leading to the deportation of the Kanakas.
The Queensland Parliament borrowed heavily from its Polynesian labourers legislation for the Aboriginals Protection and Prevention of the Sale of Opium Act of 1897, particularly the provisions dealing with basic labour practices, including the requirement for 12 month renewable agreements for Aboriginal employees, stipulating the nature and duration of work and remuneration.  The racial attitudes underlying the Act are informative because it amounted to a special law for the people of a non-European race. In other words, to borrow the terms of s51(xxvi), it constituted a law for the people of the Aboriginal race for whom the Queensland Parliament deemed it necessary to make a special law. The necessity for the law arose from: two official reports which identified quite distinct areas of concern [regarding Aborigines], particularly sexual abuses of women and children, assaults and labour exploitation, killings by Native Police forces, and “racial contamination”. 
The reports were commissioned by the Queensland Government which was seeking advice on legislation to protect Aborigines. In one report the Queensland public servant Meston advocated racial segregation to protect the non-Aboriginal population from ‘racial contamination’ and to preserve racial ‘purity’.  In the other, senior police officer Parry-Okeden advocated a pragmatic mix of intensified police and other controls over Aborigines and Torres Strait Islanders.  Ganter and Kidd note that Meston’s ‘idealism’ was repudiated in favour of Parry-Okeden’s political pragmatism.  Despite that, the Act gave enormous discretion to administrators with Aborigines under their control. It allowed, for example, Meston as an administrator scope to implement his policy of racial segregation and allowed other administrators to operate a minor business empire in the Torres Strait Islands in which they controlled the labour resource necessary for the pearling and fishing industries.  Thus the Act enabled both policies of philanthropic protection and of racial exclusion to be implemented. It also allowed for the institutionalisation of welfare:
Institutionalisation not only catered for the whole range of dominant racial concerns, it was also the usual mechanism for the state provision of welfare. In confronting a range of racial concerns, such as widespread killings, illness and destitution, and inter-racial sexual and labour relations, the 1897 Act established ‘risk’ categories which became the focus of administrative attention. 
Griffith may well have intended in 1891 to provide the proposed Commonwealth Parliament the full range of powers with his suggested race power to make laws for philanthropic protection, racial exclusion and the institutionalisation of welfare for non-Aboriginal races. The provision proposed the grant of exclusive legislative power to the Federal Parliament with respect to:
The affairs of people of any race with respect to whom it is deemed necessary to make special laws not applicable to the general community; but so that this power shall not extend to authorise legislation with respect to the aboriginal native race in Australia and the Maori race in New Zealand. 45
Griffith believed that his colleague and co-delegate from Queensland, Macrossan, would have supported the proposal. Macrossan, a representative of the northern constituencies in the Queensland Parliament, died shortly before the Convention. In his constituencies ‘the question of black labour was a burning one’.  Both Griffith and, as he believed, Macrossan supported a national race power:
because the introduction of an alien race in considerable numbers into any part of the commonwealth is a danger to the whole of the commonwealth, and upon that matter the commonwealth should speak, and the commonwealth alone. 
What Griffith had in mind was the example of the ‘immigration of coolies from British India, or any eastern people subject to civilised powers’.  He considered it necessary to provide the Commonwealth with the power because ‘no state should be allowed, because the Federal Parliament did not choose to make law on the subject, to allow the state to be flooded by such people as I have referred to’.  This concern may have arisen from Victoria’s experience with restricting Chinese immigrants with legislation in 1855 that restricted the number of Chinese that could arrive by ship. The legislation was easily evaded by Chinese arriving in South Australia and crossing the border to the Victorian goldfields.  It seems that Griffith did not intend the power to be restricted to non-European races because he adverted to the fact that some countries made special provision for European government administrators at the insistence of their European governments. He commented that:
The Dutch and English governments in the east do not allow their people to emigrate to serve in any foreign country unless there is a special law made by the people of that country protecting them, and affording special facilities for their going and coming. 
This suggests that Commonwealth legislation could provide for the needs of European administrators, and possibly business people, by providing them special conditions and amenities to encourage them to work in Australia. The later Conventions in 1897-98 held in Adelaide, Sydney and Melbourne confirmed Griffith’s original purpose for the provision.  Higgins stated that he understood it was ‘to provide for the Parliament dealing with the kanaka question’,  and O’Connor sought to have the provision clarify that it was complementary to the immigration power. 
There is debate about whether the founders intended the power to only apply to ‘alien’ races.  Quick and Garran, in their annotated commentary on the Constitution, state that the power is designed to enable laws that localise alien races within defined areas, to restrict their migration, to confine them to certain occupations and to give them special protection and ensure their return to the country from which they came.  Sawer believes their reference to ‘alien’ races did not have any precise meaning in the sense of nationality law, ‘but merely people of a “race” considered different from the Anglo-Saxon-Scottish-Welsh- Cornish-Irish-Norman (etc. etc.) mixture, derived from the United Kingdom, which formed the main Australian stock’.  Sawer relies in part for this conclusion on a comment of Barton’s at the 1898 Convention that the power was not confined to aliens, but may also apply to British subjects.  Barton may have been referring to British subjects of other (ie Asian) races (for example, Indians). Sawer’s interpretation extends the power to all races except those deriving from the United Kingdom. This seems to include ‘races’ from western Europe, including Italians, French, Germans and Greeks. His interpretation would allow laws Canadians enacted in the 19th century preventing those of the Japanese ‘race’, whether aliens, naturalised or natural born, from voting in provincial elections and the employment of Chinese in underground mines.  Canada cannot, of course, enact those laws now because of s15 of their 1982 Constitution. Sawer’s interpretation provides s51(xxvi) considerable, and concerning, scope, particularly given the extent of Australia’s present multi-racial (or multi-cultural) mix. The second most common languages spoken in Sydney, it is said, are Chinese languages. 
Kirby J expresses some doubts about the purposes intended for including s51(xxvi) in the Constitution. He rightly believes that, although Griffith may have been motivated by the blackbirding issue, the power was not necessarily intended to be limited to protecting the victims of that activity.  He adds that it is unclear whether the intention was to exclude State control over ‘alien races’ or to provide the Commonwealth with the power over aliens to deal with possible unrest and expulsion.  He observes that:
The Convention Debates, particularly those of the Melbourne Convention of 1898, show that some delegates wanted to retain power for the States, and to permit the Federal Parliament to enact, laws far from beneficial for people of minority races (such as Chinese in factories and shops, “Asiatic or African ... miner[s]” and so on). However, other delegates regarded the prospect of discriminatory legislation on the part of the new federal polity as “disgraceful” and “degrading to us and our citizenship”. 
It seems, then, that s51(xxvi) was designed, at the time of its inception, to provide the Commonwealth the power to make special laws regarding races of non- European, or at most non-British nationality. It may have also included British subjects who were of ‘non-British stock’, including those from India. The power was intended to provide for the protection of other races, whether by setting minimal labour standards or by other means. And, although it did not require segregationist legislation, it certainly contemplated it as a possibility. So although it did not necessarily seek to ensure the promotion of segregationist ideals and doctrines of racial superiority (the Commonwealth Constitution was forged from too much pragmatism to have allowed it), the provision nevertheless allowed for legislation to be based on such fundamentalist notions.
Without drawing the parallels too strongly with the Queensland legislation that may have prompted Griffith’s proposal, it is possible to envisage that s51(xxvi) was intended to grant legislative power to provide for guest workers and to establish special trade zones with differential laws applying to overseas employees in the zone. It would also have allowed laws confining racial groups to defined areas, restricting their employment and providing for their deportation. Had the exception for Aborigines not appeared in the original 1901 version of the Constitution, it would have empowered legislation that provided for the segregation of Aborigines from the rest of the population.
There is little tangible evidence as to why the founders excluded Aborigines from s51(xxvi). Thus, there is scope for speculating why they were excluded. The Chief Protector of Aborigines, Western Australia told the 1927–29 Royal Commission on the Constitution that he believed Aborigines were excluded because ‘it was widely thought that they were a dying race whose future was unimportant’.  Hanks later agreed, saying that the founders thought the Aboriginal population had been reduced to a remnant and that the delegates ‘were happy to leave to the States the responsibility of ‘dealing with’ those remnants’.  Sawer believed that the lack of s51(xxvi) debate on Aborigines revealed ‘a widespread attitude of white superiority to all coloured peoples, and ready acceptance of the view that the welfare of such people in Australia was of little importance’.  But he also believed that as the Commonwealth was not initially given any independent mainland territory and general questions of land settlement, industrial development, employment relations and education were also left to the States, ‘few of the powers given to the Commonwealth had any obvious or direct relevance to aboriginal policy’.  Galligan and Chesterman believe that the scant mention of Aborigines during the constitutional debates and in the Constitution results more from the ‘founders’ respect for the States and state powers’ than any attitude favouring or discriminating against them. 
If Harring’s analysis of early colonial law is applied to the drafting of the Constitution, the founders’ conduct cannot be seen as so benign. Indeed their scant mention of Aborigines could be seen as a positive choice to maintain their oppression. Harring observes that commentators have variously described the early Australian colonial law as ‘impotent, ambivalent, variable or standing behind the forceful dispossession of Aboriginal people while providing them the largely illusory protection of the law or providing them little thought, thus acting in a reactive way’.  But, he argues:
Once we recognise that law, in fact, existed as a powerful force in structuring colonial society in nineteenth century New South Wales, the “impotence” and “ambivalence” of law take on new meaning. Such outcomes reflect legal choices, or the choice of colonial officials not to use law in circumstances involving Aboriginal rights .... Put more crudely, legal history involves both “law” and “outlaw”, with choices to remain outside the law (or to refrain from resorting to the law) being legal choices nevertheless, the subject matter of legal history.Thus, the founders’ failure to mention Aborigines can be seen as a conscious choice for the Aborigines’ continued oppression, and not one merely motivated by respect for States rights. Not that it can be assumed the Aborigines would have been treated any more fairly by the new Commonwealth Parliament. The enactment of oppressive colonial legislation was an activity in which many founding fathers themselves had participated. 
Isaac Isaacs, himself a founding father, left no doubt about his racial attitudes when debating the first Australian electoral law in 1902 – the year after Federation. Manning Clark observes that with regard to whether Aborigines should have the right to vote: Isaac Isaacs argued that until such time as Aborigines were thought worthy to vote for state parliaments members of the Commonwealth Parliament should not consider them worthy to vote in a federal election. The Aborigines, he maintained, did not have the intelligence, interest or capacity to enable them to stand on the same platform with the rest of the people of Australia. No one demurred.  Isaacs’ views were not idiosyncratic.  He represented the attitudes of his time. And they were times of heightened, and to an extent unprecedented, racial bigotry. Attwood notes, for example, that attitudes to Aborigines in the bush were relatively harmonious before the late 19th century, because Aboriginal labour was valued by the pastoralists.  But they began to change as ‘rural communities which increasingly came to be comprised of newcomers, women now as well as men, who were relatively unfamiliar with Aborigines and who increasingly prized personal and civic respectability’.  They harboured deep fears and suspicions of the Aborigines which prompted the discriminatory legislation enacted around the turn of the century. Racial fears and bigotry were not confined to the Aborigines, the Chinese in particular were also to fall victim. They began to arrive in Australia in the 1840s as cheap labour for the pastoral industry and were followed by tens of thousands more when gold was discovered.  Resentments grew on the goldfields where the ‘Eureka flag flew as the miners attacked the Chinese; radicalism and nationalism had a racist element’.  Kercher believes that:
White attitudes were based on greed, a deep feeling of their own moral superiority and fear of what the Chinese might bring, the spread of vices such as opium smoking and gambling, racial conflict and cheap competition for white workers .... Australian egalitarianism was based on an equality of white men alone. Like Aborigines, Asians were left out of the embrace of mateship. Such icons as The Bulletin magazine and the Labor party, and most politicians, openly favoured immigration restrictions based on race at the end of the nineteenth century.
Interestingly, in the 1860s conservatives supported the British Empire’s open-door immigration policy and opposed anti-Chinese laws, but were accused by liberals of doing so for cheap labour on their properties.  But their support eroded, and increasingly restrictive laws were introduced against the Chinese, despite objections from London about their racist character.  Higher poll taxes and greater restrictions on the numbers of Chinese allowed on each arriving ship were imposed. The increasingly racist content of Australian colonial legislation was by 1896 concerning London to the extent that Secretary of State for the Colonies, Joseph Chamberlain, informed the premiers that while he ‘was sympathetic to the colonies’ aims and to the concern about the threat to white workers, he was worried that the openly racial basis of the legislation offended Asian members of the empire as well as Britain’s ally, Japan’.  He proposed a more covert means for achieving the same ends – the language test. This infamous device, which lasted until 1958, was used to exclude entry into Australia of those the government considered undesirable or inconvenient. Even those fluent in English were excluded if their race, political belief or some other characteristic did not suit the government of the day because the language test could be set in any European language. 
Deep springs fed the racial attitudes of the late 19th century, and their widely divergent sources led to complex and sometimes contradictory racial attitudes. From one spring flowed the influence of the British humanitarian movement, which had gained impressive victories, including the abolition of British slavery, during the early part of the century.  From another flowed economic self interest. Pastoralists, gold miners and rural (and to a lesser extent, urban) workers, as we have seen, believed it was in their economic interest to despise the Chinese and other races. From yet another source sprang the influence of the pseudo-sciences of social-Darwinism and eugenics, which fed deeply harboured fears and racial arrogance. Fears existed of being outnumbered and contaminated by other races, including the Chinese, Indian and Kanaka people. This, of course, presumed a separateness of the ‘European’ races from the other feared races. Co-existing with fear was racial arrogance born of the assumption that European races were superior to others. This arrogance led to bureaucratic control over almost all aspects of daily Aboriginal life and the belief amongst some that protection was required to smooth the pillow for a dying race. An (often unarticulated) undercurrent flowing through many of these sentiments and policies was the eugenic notion of racial strength and purity. 
The eugenic movement was not particularly well organised or evident in Australia, at least not to the extent it was in the United States, Britain and other parts of Europe. Garton, however, warns against underestimating its influence on Australian ideas and attitudes,  which were informed by European and American debates on race and biology. Eugenics was founded by Francis Galton, who coined the term in 1883, and who took inspiration from his cousin Charles Darwin’s book On the Origin of Species (published in 1859).  He later defined eugenics as ‘the study of agencies under social control that may improve or impair the racial qualities of future generations either physically or mentally’.  Galton was not alone in attaching significance to race, as Hannaford explains:
During the period from 1890 to 1915, race as an organizing idea claimed precedence over all previous formulations of nation and state. Although the works of many racialist writers of the period are virtually unreadable today ... they attracted vast audiences in Germany, France, Britain and the United States, who were greatly excited by racial ideas.
Eugenics gained a growing number of adherents in the late 19th century, which developed into a flood of popularity by the early 20th. Galton was knighted in 1907 and his eugenic ideas were exalted in the popular press. His adherents included Alexander Graham Bell, George Bernard Shaw, Harold Laski and Beatrice and Sidney Webb.  It was seen by many as ‘a promising new instrument for the release of civilization from the uncertainties and contingencies of existing politics’.Tragically, and in a way its adherents had not counted on, it achieved that end. Although eugenics was primarily interested in biological progression and ‘improvement’ by weeding out weaker biological strains in the human species, racial concerns readily fell into the vortex of eugenic interest and concern. The eugenicists, like many others who subscribed to various other racial theories, were interested in the traits of different ‘races’ – hair and skin colour and texture, facial shape, the length of limbs, and so on. And of course they were interested in measuring and comparing racial traits of intelligence. In the pursuit of scientific endeavour, eminent scientists derived reams of scientific data that confirmed their own prejudices, as well as that of the general populace.  Amongst the leading eugenic scientists was the American, Davenport, who in 1911 identified the hereditary incidence of Huntington’s chorea, haemophilia, albinism and polydactylism.  He proceeded to consider whether state laws limiting the entitlement to marry within and between families had a sound scientific basis. He recommended the better administration of the laws by eugenics boards staffed by biologists who would issue marriage licences on the basis of scientific data. His recommendation was not followed in the US, but was given effect by the Nazi law
of 28 June, 1933.  Meanwhile a number of United States states introduced sterilisation laws which were aimed at the feebleminded, to prevent them breeding children of inferior intelligence.
These laws were at times successfully challenged in the courts until the Supreme Court decision of Buck v Bell  in which the Court upheld the constitutionality of Virginia’s sterilisation law on the basis that the law was within the police power of the state, it provided due process of law and did not constitute cruel or unusual punishment. This allowed the authorities to sterilise Carrie Buck, who after giving birth to a child out of wedlock, was tested as having the IQ of 9 years (and her mother tested at 8 years). In giving the Court’s opinion, Justice Oliver Wendell Holmes stated that:
We have seen more than once that the public welfare may call upon the best citizens for their lives. It would be strange if it could not call upon those who already sap the strength of the State for these lesser sacrifices ... in order to prevent our being swamped with incompetence .... 
The decision gave added respectability to the eugenic cause. About half the United States states legislated to allow the sterilisation of those in prisons and other institutions on the basis of their feeblemindedness, social inadequacy and retardation.  ‘By the mid-1930’s some 20,000 sterilisations had been legally performed in the US’.  These were not limited to those residing in institutions. A former member of the Montgomery County, Virginia, Board of Supervisors recalls state sterilisation authorities raiding whole families of ‘misfit’ mountaineers in the 1930s.  Hitler’s sterilisation law was introduced in 1933 and required the sterilisation of those suffering ‘feeblemindedness, schizophrenia, epilepsy, blindness, severe drug or alcohol addiction and gross physical deformity’.  The law met with international objection, to which the Nazis responded by pointing to the United States model they had adapted. They enforced the new law, however, with a vigour and efficiency that far exceeded the US. Although initially the program was not based on racial categories, it was not long before this changed. Sterilisation was extended to include those with infectious diseases or different ‘racial’ backgrounds. In 1939 the law moved beyond sterilisation to euthanasia of certain classes of the mentally diseased or disabled in asylums. Amongst the classes were Jews, to whom the procedure applied, regardless of their mental health. 
This is not to suggest that eugenic theories on social biology inevitably lead to racial segregation and racial genocide. The United States experience confirms this. And indeed many eugenicists either did not support the more extreme racial theories or gave it lukewarm support. Searle comments that:
The contention that British eugenists were not ‘super-nationals’ or glorifiers of war might be countered with the objection that they were nearly all of them pronounced racialists, as their very language proves. Phrases like ‘the traditions of the race’, ‘racial instinct’ and ‘race-regeneration’ occur with monotonous regularity in eugenical literature. This by itself, however, is not conclusive, since these phrases were also regularly employed by contemporaries who cannot by any stretch of the imagination be called ‘racialist’. In the early twentieth century the word race seems to have been interchangeable with ‘nation’, ‘community’, or even ‘people’.Thus, the term ‘race’ was interchangeable with other terms dealing with group identity, which themselves related to different ideologies. Many eugenicists, particularly the British, did not believe in the virtues of racial purity, and to the extent that they did, they were not necessarily anti-Semitic. Indeed, some believed that the alleged racial purity of the Jews and their above average intelligence confirmed the virtues of racial homogeneity. In general, however, there was a belief in racial hierarchy, with all the consequent prejudice that it confirmed or engendered.
Racial eugenics and other theories of racial hierarchy suffered a loss of credibility and acceptability after the second world war because of their adoption (and to an extent) distortion by the Nazis. Their theories, although extreme, were sourced from the thinking and politics of the late 19th and early 20th century. On his own admission, Hitler obtained all of his ideas confirming the singular importance of race from the period immediately preceding 1910.  The Nazi ideas and attitudes derived in part from a misreading of Nietzsche (who ironically saw the Jews as an example of a superior race),  the formation of the antisemitic movement in Germany, Austria-Hungary and France in the 1880s, partly in response to the influx of Jews from the Russian pogroms, and the popularising of myths on ‘the jewish conspiracy’ and Aryan supremacy that were pedalled by Drumont in France.  Extremist anti-Semitic ideology and politics were not invented by the Nazis, as the late 19th century anti-Semitic movement and the extremist rantings of the Mayor of Vienna, Karl Lueger (which inspired Hitler) attest. Nor did they invent eugenic science and the idea of Aryan racial superiority, which, ironically, derived from France. But the Nazis combined these extant theories and policies to appalling effect. They crushed the delicate liberal notions of due process, equality before the law and the rule of law. And as Kirby J points out, this was not in one fatal blow, but by means of attrition, so the early laws were capable of judicial assent. 
Although eugenic theories on social biology do not inevitably lead to racial segregation and racial genocide, the United States experience of sterilisation laws alone highlights the dangers that arise if the courts fail to maintain an intense scepticism of laws that single out vulnerable groups in society for the deprivation of basic human rights. As Hannaford reminds us:
However well-intentioned these scientific societies and journals may have been at the time of their formation, and however much we may wish to distance their illustrious academic and industrial founders from what happened in the death camps, it has to be remembered that the first experiments on the feebleminded were sanctioned by states in the United States and justified by the United States Supreme Court. 
The European and American racial theories of the late 19th century, it can be reasonably supposed, offered succour to those who believed in racial segregation, which in turn played a role in inspiring legislation applying to the Chinese, Kanakas and Aborigines. The Queensland public servant Meston, for example, as we have seen, wrote a report advocating racial segregation to protect the non- Aboriginal population from ‘racial contamination’ and to preserve racial ‘purity’. Labelling someone like Meston a eugenicist is, however, potentially misleading. Most, like him, fed their attitudes from streams that were sourced from humanitarian, neo-Darwinist, eugenic, Christian and other well-springs of belief and ideology. And each stream of thought carried differing quantities of the ideologies sourced from the springs.  And although the Constitutional Conventions were attended by men ‘who were in general sensitive, humane, and conscious of religious and social duties to the less fortunate sections of the community’, 112 they nevertheless revealed in their debates ‘a widespread attitude of white superiority to all coloured peoples’. 
During the period after World War II the community’s attitudes and assumptions about race began to undergo a profound change. The change was not immediate and did not develop according to a consistent pattern, but it was profoundly influenced by the horrors of the Nazi race program. The change in general attitude is marked by the 1967 amendment to s51(xxvi), which was motivated by a general desire to enable the Commonwealth to make laws which treated Aborigines as full and equal citizens. The amendment was prompted by the realisation that the States’ exercise of power over Aborigines was a failure. Ironically, the failure was in part due to segregationist legislation motivated by the same assumptions about race that created s51(xxvi).
On its face the changes made by the 1967 amendment were ‘minimalist’, as Gaudron J describes it. 
The original words of the relevant parts of s51 remained unaltered except for the removal of eight words shown here in italics:
Parliament shall have the power to make laws for the peace, order and good government of the Commonwealth with respect to: (xxvi) The people of any race, other than the aboriginal race in any State, for whom it is deemed necessary to make special laws. 115
A narrow literalist reading of the Constitutional alteration suggests that the amendment was designed merely to extend the Commonwealth’s pre-existing racial power (to discriminate against racial groups) to Aborigines.  Gummow and Hayne JJ suggest that the pre-amendment scope of the provision remains essentially unaltered, except for the extension of powers over Aborigines, because the amendment was driven by considerations of federalism:
The treatment in the ‘yes’ case of the proposed alteration to the power of the Commonwealth legislature emphasised considerations of federalism. It did not speak of other limitations upon the nature of the special laws beyond confirming that they might apply to the people of the Aboriginal race ‘wherever they may live’ rather than be limited to the Territories.  It is true that the Yes Case made it clear the amendment would not provide the Commonwealth with an exclusive power over Aborigines and that ‘the Commonwealth’s object will be to co-operate with the States’,  but the statements appear to be aimed at ameliorating concerns about the undue extension of Commonwealth power rather than stating an end itself. The substantive purpose, as the Yes Case makes clear, was for the Commonwealth and the States ‘to ensure that together we act in the best interests of the Aboriginal people of Australia’. 
The amendment was about Aborigines, not federalism. The amendment of s51(xxvi) occurred when there was a general sense in the Australian community that denying Aborigines ‘full citizenship’ and equality under the law was unjust, and continuing the injustice would adversely affect Australia’s international reputation. Post-war reaction to the grotesqueness of the Nazi program of racial extermination led to a change in international attitudes toward race and the eugenic and social biological theories of racial superiority that had underscored it. The Nazis, as we have seen, took abstract theorising and virulent racist propaganda to their (il)logical extreme by putting words into action with a program of racial purification which as we know involved the mass killings of millions of Jews, Gypsies and Poles. The Nazis’ activities brutally illustrated to the post-war world the extremes to which attitudes of racial separation and superiority could be taken. This in turn led to the creation of institutions, policies and laws in the post-war period that promoted racial tolerance and equality. The 1967 amendment to the Australian Constitution was a child of the post-war ambitions to promote racial harmony.
The antidotes for theories of racial superiority became racial tolerance and equality. The United Nations was established, largely to prevent the outbreak of international conflicts and to promote the interests of the international community. One of its most significant foundation documents was the Universal Declaration of Human Rights 1948. This, and the International Convention on the Elimination of all forms of Racial Discrimination 1965 (the ‘Convention on Racial Discrimination’), helped lay the foundations for national laws and constitutions to provide for racial tolerance and equality. Australia became a signatory to the Convention on Racial Discrimination on 13 October 1966 and ratified it on 30 September 1975, which led eventually to the enactment of the Racial Discrimination Act in 1975 which gave domestic effect to the Convention.
The Convention on Racial Discrimination was preceded by a number of general human rights statutes and conventions including the European Convention for the Protection of Human Rights and Fundamental Freedoms which was drawn up by the Council of Europe in 1950 and came into force in September 1953.  Article 14 of the Convention prohibits discrimination, including discrimination on the grounds of race, in the enjoyment of fundamental rights and freedoms. Canada enacted its Bill of Rights in 1960, which was an ordinary federal statute dealing with the protection of a wide range of fundamental rights, including racial equality. Section 1 recognised the right of the individual to equality before the law and the protection of the law regardless of the individual’s race, national origin, colour, religion or sex. The United States’ Supreme Court changed its understanding of the Fourteenth amendment to its Constitution which provided that all its citizens were entitled to equal protection of the laws. It overturned the ‘separate but equal’ doctrine of Plessy v Ferguson  (which upheld a Louisiana statute that required blacks to ride in separate train carriages on the basis that the train facilities were separate but equal) in favour of the removal of racial segregation in the 1954 decision of Brown v Board of Education.  The legislative and judicial changes represented, and at times led, social attitudes toward race. The pattern of change in attitudes was at times traumatic, as illustrated by social unrest in the US, particularly during the 1950s and 1960s, and South Africa in the 1970s to 1990s. The pattern of change was rarely even, and no doubt in places superficial. The race debates in Australia in response to the Mabo (No 2)  and Wik  High Court decisions reveal how superficial the mask of tolerance can be.
By the mid-1960s, then, there was a growing awareness in the world community, including Australia, of the dangers, irrationality and injustice that racial segregation and discrimination can cause. This placed the race power in s51(xxvi) of the Constitution in a different light, which affected the meaning of the term ‘race’ in the provision. So although the spelling of the term had not altered in the 66 years since Federation, its meaning and value had. By the end of the 19th century ‘race’ was the term used for dubious biological and historical categorisations for the purpose of asserting hierarchies of ability, intelligence and social worth. This in turn justified segregating people and depriving them of their basic rights and entitlements. By the 1960s fears of being overwhelmed by ‘alien’ (ie, Asian) races had subsided in Australia, and the consequences of taking ideas of racial superiority to their extremes had been illustrated by the Nazis. According to Kirby J:
The laws of Germany and South Africa ... provide part of the context in which par (xxvi) is now understood by Australians and should be construed by this Court. I do not accept that in late twentieth century Australia that paragraph supports detrimental and adversely discriminatory laws when the provision is read against the history of racism during this century and the 1967 referendum in Australia intended to address that history. When they voted in that referendum, the electors of this country were generally aware of that history. They knew the defects in past Australian laws and policies. And they would have known that the offensive legal regimes in Germany during the Third Reich and South Africa under apartheid were not the laws of uncivilised countries. Both in Germany and in South Africa the special laws enacted would probably have been regarded as unthinkable but a decade before they were made. They stand as a warning to us in the elaboration of our Constitution. 
What then was the purpose and effect of the constitutional amendment? The mischief rule can assist with an answer. Applying the rule involves comparing s51(xxvi)’s post-amendment meaning with its original purpose to discover the mischief with which the amendment sought to deal. The essential question, then, is why did Parliament bother asking the Australian electors to agree to the provision’s amendment? Constitutional changes are not lightly embarked upon. The process is expensive and Australia’s experience of constitutional amendment reveals the electorate’s resistance to change. Consequently, there must have been a perceived problem that could only be dealt with by the grant of Commonwealth power.
The perceived problem, it is submitted, was the legislation and policies of at least some States that discriminated unfairly against Aborigines and which denied them full participation in the life of the nation as equal citizens. Legislation applying to Aborigines in some States were segregationist. Queensland’s Aboriginals Protection and Restriction of the Sale of Opium Act 1897, for example, provided for the removal of Aborigines and the maintenance of their order and discipline, the custody and education of their children and their employment and apprenticeship.  These matters were dealt with in a way that allowed extreme levels of administrative discretion and control, that did not apply (or would not have been tolerated) by the majority non-Aboriginal population. The Act also provided for the prohibition of Aboriginal rites and customs, and allowed administrators to control the marriage of female Aborigines and to prohibit visitors to Aboriginal camps.  In 1934 the Act was amended to prohibit sexual intercourse between an Aboriginal woman and a non-Aboriginal man and to allow administrators to cancel written agreements to employ an Aboriginal or a ‘halfcaste’.  Wills and deeds of gift or transfer of land were declared invalid unless approved and witnessed by the Chief Protector for Aborigines. The Minister could declare an Aborigine or ‘half-caste’ to be uncontrollable and to be kept in an institution (including a prison) without any rights of administrative or judicial review. In 1939 the Act provided for the removal and detention of any Aborigine in a reserve.  The Act remained on the statute books until its repeal in 1965.
It was recognised by some not long after Federation that the States were treating Aborigines poorly and that Commonwealth power and responsibility were required to deal with the situation. In 1911 a non-government conference called for full federal responsibility for Aborigines and in 1929 a minority report of the Royal Commission on the Constitution recommended federal financial assistance to States with the most numbers of Aborigines.  In 1936 federal control over Aborigines was raised but rejected, with an agreement that there be regular meetings of State and Federal officials regarding Aboriginal affairs. And in 1944 Prime Minister Curtin proposed a referendum (which was defeated) that, amongst other things, proposed that the Commonwealth have a power with respect to ‘the people of the Aboriginal race’.
Concern that the Commonwealth should be doing something to assist Aborigines mounted in the post-war period, but s51(xxvi) was perceived as a bar to federal laws for Aborigines.  In 1958 the Joint Parliamentary Committee on Constitutional Review was established to consider constitutional reform.  It recommended the repeal of s127. Concern also surfaced in the 1950s and early 1960s about the denial to the Aborigines of the right to vote in Queensland and Western Australia. Federal law only allowed Aborigines and Torres Strait Islanders who could vote in State elections to vote in federal elections. As a result, the Federal Electoral Act was amended in 1962 to allow all Aborigines and Torres Strait Islanders to vote in federal elections despite State laws. Western Australia amended its Electoral Act that year to allow Aborigines the vote.  Queensland did not allow it until 1965, and even then subject to the proviso that anyone ‘who influences or attempts to influence in any manner or by any means whatsoever’ an Aboriginal person to enrol commits an offence.  The rising concern about Aborigines also resulted in a number of federal parliamentary motions and bills (in 1962, 1964, 1965 and 1966) each requiring a referendum for altering the Constitution with regard to Aborigines.
The general mood preceding the referendum is partially captured by Professor Sawer, who wrote in 1966 that:
Today there is a public conscience concerning the aborigines; since about 1956, steps have been taken to give them increased citizenship rights and liberties, to atone for wrongs done them by the white man, and to secure their full participation in government ....
The urge for constitutional change was for more than simply granting federal powers over Aborigines. There was a more general sense that Aborigines had been badly treated and should be treated as equal citizens. A successful bill was introduced by Prime Minister Holt on 1 March 1967 leading on 27 May 1967 to Australia’s largest yes vote for constitutional change. 
The electors were asked to approve the amendment of s51(xxvi) and the deletion of s127, which prohibited the counting of Aborigines when reckoning the population. In the official Yes Case provided to electors before the referendum it was stated that the purposes of the proposed amendments were ‘to remove any ground for the belief that, as at present worded, the Constitution discriminates in some ways against people of the Aboriginal race, and, at the same time, to make it possible for the Commonwealth Parliament to make special laws for the people of the Aboriginal race’.  The Yes Case later repeated these propositions:
The proposed alteration of this section [ie, 51(xxvi)] will do two things. First, it will remove words from the Constitution that many people think are discriminatory against the Aboriginal people.
Second, it will make it possible for the Commonwealth parliament to make special laws for the people of the Aboriginal race, wherever they may live, if the Parliament considers it necessary. 
The Yes Case continued by questioning, after quoting s127, why it was included in the Constitution:
Well, there were serious practical difficulties in counting the Aborigines in those days. They were dispersed, and nomadic. Communications in inland Australia were poor, and frequently non-existent. Today the situation is very different and counting is practicable.
Our personal sense of justice, our commonsense and our international reputation to a world in which racial issues are being highlighted every day, require that we get rid of this out-moded provision.
Its modern absurdity is made clear when we point out that for some years now Aborigines have been entitled to enrol for, and vote at, Federal Elections. Yet s127 prevents them from being reckoned as “people’ for the purpose of calculating our population, even for electoral purposes.
The overwhelming and record breaking yes vote for the referendum proposal was influenced in part by Parliament’s unity and argument for the need for reform, and media stories and images which alerted the public to the poverty and degradation suffered by indigenous people, who at the time were living in one of the world’s richest nations. As Brennan J observed in the Tasmanian Dam Case, the 1967 referendum was an ‘affirmation of the will of the Australian people that the odious policies of oppression and neglect of Aboriginal citizens were to be at an end, and that the primary object of the power is beneficial’.  And as Sawer noted in 1966, there was a desire to ‘atone for wrongs done them by the white man’ and to ensure that Aborigines could fully participate in government.  In the Parliamentary debates for the bill to amend the Constitution  repeated mention was made of a need to enable Commonwealth laws for the benefit of Aborigines. Prime Minister Holt said that the bill was introduced in response to the popular impression that the words excluding Aborigines in s51(xxvi) were discriminatory.  And the object was to enable the Commonwealth to work with the States for the advancement of Aborigines.  The opposition agreed. 
The intention therefore was to enable the Commonwealth (possibly in cooperation with the States) to make laws for Aborigines, including by removing the bar to the entitlements of citizenship that effectively had been denied them by some States. The mischief the alteration was designed to address was the State laws that discriminated against Aborigines and denied them the basic fundamental rights and freedoms enjoyed by other non-Aboriginal members of the community.  The mischief was also laws that were based on the value that some races were more equal than others because of inherent ‘racial characteristics’ that were superior to others.
Although the amendment was motivated by the desire to improve the position of Aboriginal people, the desire was based on assumptions about the meaning and value of race which were substantially different to those held at Federation. Section 51(xxvi) was infused, by its amendment, with values and meaning which differed from, and to an extent contradicted, those held at Federation. Thus, the challenge, if not the obligation, on the interpreter is to interpret the provision in a way that maintains fidelity to the text while accommodating the meaning and values imported into the provision by its amendment.
A law purporting to be exercised under s51(xxvi) obviously must be a law regarding a race of people. What may not be immediately obvious is that it cannot be assumed that a law’s categorisation of a race will be constitutionally valid. The matter is open to judicial review. To illustrate the point, it is highly debatable whether Parliament can enact a law for the ‘Dutch race’. But deciding whether Parliament has validly categorised a group of people as a race requires untangling the knot of unarticulated and articulated legislative assumptions about the group’s alleged distinct racial characteristics. Similarly, the Constitution’s assumptions about the term ‘race’ need to be made explicit. If there is harmony between the legislative and constitutional assumptions about the relevant characteristics for defining a racial group, the challenged legislation’s racial categorisation will likely be valid.
A clear understanding of the term ‘race’ is also important because the characteristics that are perceived to exist for defining a group of people as a distinct race arguably bear a relationship to the powers that can be exercised regarding the group. Thus the question of whether a group is properly defined as a race in challenged legislation is a preliminary one in deciding the validity of its racial categorisation. Identifying the rationale for categorising a group as a race can also assist in determining whether the legislation properly exercises power with regard to the race.
As has been discussed earlier in Part 2, the categorisation of a group of people as a race is often a contentious one, and not one that can be done on the basis of ‘objective’ scientific evidence. Rather it is now seen to be socially determined, which means that racial categorisation can vary in time and between places. Racial categorisation involves selecting features or characteristics of a group which are seen to be unique to the group. The characteristics which are perceived to be relevant in defining the group as a distinct race can vary over time and place. The characteristics which were considered relevant for defining Aborigines as a distinct racial group in 1901 differ from those that will be considered relevant a century later. In 1901 the grounds for racial distinction were based on pseudobiological or nationalist conceptions of race, which led to the assumption, for example, that Aborigines were a distinct and inferior sub-species or group and that their difference in appearance was linked to their sub-species or sub-social status. As McGregor observes, in the literature of classical evolutionism, which was in vogue at the turn of the century:
[T]he concepts of race and progress were both pressed into service to fill a major hiatus in the narrative account of the origins of man and his civilisation. The various races were rendered stages in a developmental sequence; but unlike the superficially similar scheme of the Enlightenment, in the evolutionary version it was a developmental sequence not only of society but also of biology. Human anatomy, mentality, culture and society all marched in step, within discrete racial units. By making the transition from savagery to civilisation an integral part of the evolution of the human species, the latter process was imbued with both meaning and direction. Charles Darwin, himself a liberal humanitarian who was opposed to Negro slavery, 147 held an ‘optimistic view’ of the evolution of humans. His views were enormously influential by the end of the 19th century. He believed that humans were gradually evolving from primitives to a superior race, and this process was a continuing one, so that:
At some future period, not very distant as measured by centuries, the civilised races of man will almost certainly exterminate and replace throughout the world the savage races. At the same time the anthropomorphous apes ... will no doubt be exterminated. The break will then be rendered wider, for it will intervene between man in a more civilised state, as we may hope, than the Caucasian, and some ape as low as a baboon, instead of as at present between the negro or Australian and the gorilla. 
Darwin was only one of numerous influential scientists and commentators who ranked human races according to their level of development, and the Aborigines invariably ranked at the bottom.  Thus, the characteristics that were seen to distinguish Aborigines as a racial group were their lower forms of social behaviour and biological development. These identified characteristics served as the basis for legislative attention. Legislation was required to separate many Aborigines from the rest of the community, to protect them from the rest of the community and vice versa. It can be recalled from the discussion in Part 4 that the Queensland Aboriginals Protection and Prevention of the Sale of Opium Act of 1897 sought to protect Aborigines from abuse by whites and (in the view of some) to protect whites from racial contamination. The lower status of Aborigines made them vulnerable to abuse, requiring legislative intervention. But the legislation that protected them also assumed that they were not capable of, or entitled to, the rights and privileges of other citizens because of their lower social and biological status, and so denied them the entitlements of full and equal citizenship.
Racial categorisation of Aborigines on the basis of social and biological status became increasingly problematic with intermarriage. The response was to subcategorise Aborigines in order to retain the status ranking of racial groups. As Hanks notes, the genealogical approach to defining Aborigines persisted from Federation at least until the 1967 referendum. He provides the example of the 1966 census which asked respondents to nominate their race and explained, ‘If of more than one race give particulars, for example,  Chinese ...’  The States continued to define Aborigines by reference to proportions of ‘Aboriginal blood’. Queensland and Western Australian legislation of the mid-1960s defined an Aborigine as ‘a person who has a preponderance of the blood of an Aborigine.’  The attitude at the federal level soon began to show signs of change, so that the 1971 census asked for a person’s ‘racial origin’, but ‘If of mixed origin, indicate the one to which he considers himself to belong’. 
Today the basis for distinction is established more on sociological grounds. In Mandla v Dowell Lee  , for example, the House of Lords was required to consider whether Sikhs were a ‘racial group’ for the purposes of the Race Relations Act 1976 (UK). The Act allowed an ‘ethnic group’ to fall within the Act’s definition of race. Lord Fraser noted that:
It would be absurd to suppose that Parliament can have intended that membership of a particular racial group should depend on scientific proof that a person possessed the relevant distinctive biological characteristics (assuming that such characteristics exist) ... the briefest glance at the evidence in this case is enough to show that, within the human race, there are very few, if any, distinctions which are scientifically recognised as racial. 
He proposed a number of conditions he considered essential for defining an ethnic group, including a long shared history, a cultural tradition of its own, a common geographical origin or descent, a common language, a common literature, a common (and different) religion and being a minority in the general community.  Brennan J makes mention of similar categories in the Tasmanian Dam Case. He says that the biological element is essential to determining who was an ‘Aborigine’, but adds that:
It does not ordinarily exhaust the characteristics of a racial group. Physical similarities, and a common history, a common religion or spiritual beliefs and a common culture are factors that tend to create a sense of identity among members of a race and to which others have regard in identifying people as members of a race. In the same case Deane J defines Aborigines according to what he considers to be the conventional meaning of the term, namely, ‘a person of Aboriginal descent, albeit mixed, who identifies himself as such and who is recognised by the Aboriginal community as an Aboriginal’. The characteristics seen as defining the Aboriginal race in 1901 differ markedly from those identified by the above-mentioned courts in the 1980s. The assumptions underlying those categorisations have also markedly changed. In 1901 the underlying assumption was of racial hierarchy, whereas in the 1980s it was of self-identity largely based on unique social characteristics that cannot necessarily be ranked according to hierarchy. Identifying and making explicit the underlying assumptions for racial categorisation offers a way of understanding some of the processes which lead the legislature to enact race based legislation. If a legislature assumes that a racial group is identifiable by its inferior characteristics, then it is not surprising if it enacts legislation that confirms the group’s inferior status and denies it entitlements due to non-members of the group. If a racial group is characterised as being equal in status, but different in terms of certain social norms, then legislation would rationally treat it as distinct and equal. Thus it can be generalised that the process for deciding whether and why a group is considered to be a distinct race is relevant to the question (put either in a political or constitutional context) of whether the legislation is appropriate. That is, why we consider a group to be a distinct race relates to what we can and should legislatively do about it.
The relationship between the characteristics that define a racial group and the power that may be exercised with regard to the group has arisen in consideration of the scope of s51(xxvi) of the Constitution. Stephen J notes in Koowarta v Bjelke-Petersen,  for example, that:
Because the reference to ‘The people of any race’ is qualified by the requirement that they should be such that it is deemed necessary to make special laws for them, they must possess some quality which calls for laws special to themselves. This requirement is more than a mere qualification of the power; it also predicates a character which laws made under par. (xxvi) must possess: they must be special laws, in the sense of having some special connexion with people of any race .... Although it is people of ‘any’ race that are referred to, I regard the reference to special laws as confining what may be enacted under this paragraph to laws which are of their nature special to the people of a particular race. It must be because of their special needs or because of the special threat or problem which they present that the necessity for the law arises; without this particular necessity as the occasion for the law, it will not be a special law such as s51(xxvi) speaks of. 
In other words, there must be some distinct quality (or characteristic) of the racial group which exists before the need arises for a special law regarding that distinct quality. The law may deal with the need, threat or problem that relates to the racial quality. For example, it may be recognised that a particular community or group of Aborigines form a racial group because they hold a distinct body of traditions, observances, customs and beliefs regarding particular areas, objects or relationships.  On that basis a law (the Heritage Protection Act, for example) may be enacted to protect those areas, objects or relationships from desecration.
The Native Title Act Case appears, however, not to support Stephen J’s reliance on the term “special” to qualify the scope of s51(xxvi). It was stated in that case that ‘‘‘[s]pecial” qualifies “law” [and] does not relate to necessity’ with the consequence that the ‘special quality’ of the law in question is to be ‘ascertained by reference to its differential operation upon the people of a particular race’  . Thus, a special law appears to be merely a law that treats one race differently from others. Stephen J’s general point, however, is that s51(xxvi) limits its scope to laws that relate to the distinct needs, threats or problems relating to the race in question. This point is picked up by Gaudron J in the Hindmarsh Island Bridge Case, although she relies on the term ‘deemed necessary’ rather than ‘special laws’ for creating the relationship between the characteristics that distinguish a race and the legislative mechanisms for dealing with the distinct characteristics. She states that:
The criterion for the exercise of power under s51(xxvi) is that it be deemed necessary – not expedient or appropriate – to make a law which provides differently for the people of a particular race or, if it is a law of general application, one which deals with something of ‘special significance or importance to the people of [that] particular race’. Clearly, it is for the Parliament to deem it necessary to make a law of that kind. To form a view as to that necessity, however, there must be some difference pertaining to the people of the race involved or their circumstances or, at least, some material upon which the Parliament might reasonably form a political judgment that there is a difference of that kind. Were it otherwise, the words ‘for whom it is deemed necessary to make special laws’ would have no operation and s51(xxvi) would simply be a power to make laws for the people of any race.
Parliament must therefore make its decision to enact a special law on a reasonable basis. There must be a difference between the group to whom the law will apply and the rest of the community which is appropriate for categorising them as a racial group—that is, a ‘relevant difference’ must exist.  Two consequences follow, according to Gaudron J, once the racial differences are identified: first, a valid law can only relate to the matters of difference,  and second, the law must be reasonably capable of being viewed as appropriate and adapted to the difference asserted.  The second consequence is that there be a rational connection between the racial difference and the legislative measures applying to it.
The requirement that the legislature make rational judgments regarding racial laws has an antecedent in the equal protection clause (the Fourteenth Amendment) under the United States Constitution. The clause was introduced after the Civil War and was aimed at providing emancipated blacks full and equal status. The Supreme Court’s interpretation of the provision initially did not remove racially discriminatory laws.  This interpretation, however, was overturned in the landmark decision of Brown v Board of Education  in which Warren CJ found that the requirement that children be taught in segregated schools, even with equal physical facilities, ‘generates a feeling of inferiority [amongst blacks] as to their status in the community that may affect their hearts and minds in a way unlikely ever to be undone’.  The decision did not lead the Supreme Court to subsequently overrule statutes which made distinctions amongst people leading to different legal treatment. Indeed it was recognised that there are perfectly fair and rational grounds for treating categories of citizens differently.  For example, refusing to allow under 16 year olds to apply for a driver’s licence may be fairly done to protect road users.
Under the United States Constitution legislation can fairly classify people for unequal treatment if the categorisation is rational, relevant to achieving a legitimate end and made to achieve a legitimate legislative purpose.  Legislation should be rational in the sense that like groups should usually be treated alike; be relevant in that the categorisation must be necessary for achieving the legislative purpose; and have a purpose legitimate in that it must be for the public interest or deal with an issue which the legislature is permitted to pursue.  The Supreme Court uses a two tiered scrutiny test: strict and minimal. Of interest for present purposes is that the Court considers legislative categorisation on the basis of race, religion and nationality as inherently suspect, requiring strict scrutiny.  The Court considers that blacks and whites have few inherent differences and therefore presumes that legislation which employs racial categories is likely to be making racial distinctions for irrational purposes which do not further legitimate legislative ends.  Legislation which categorises groups on an economic and social basis is not inherently suspect, however, and is therefore subjected to minimal scrutiny. Here the onus is on the challenger to prove that the classification of a group for discriminatory treatment lacks a rational relationship to the legislative object. 
Similar concerns about race laws are reflected in s15 of the Canadian Charter of Rights and Freedoms 1982, which states that:
Every individual is equal before and under the law and has the right to the equal protection and equal benefit of the law without discrimination and, in particular, without discrimination based on race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.
Section 15(2) specifically allows programs or activities that have the object of ameliorating disadvantage. The present tests for s15 were decided in a trilogy of 1995 Supreme Court cases: Miron v Trudel,  Egan v Canada  and Thibaudeau v Canada  . It was decided that the analysis under s15(1) involves two steps. First the complainant must show a denial of equal protection or benefit of the law compared to other people. That is, whether a legislative distinction ‘has the effect on the claimant of imposing a burden, obligation or disadvantage not imposed on others or of withholding or limiting access to benefits or advantages which are available to others’.  Second, the law must be shown to be discriminatory by denying one of the equality grounds mentioned in s15 and that the unequal treatment is based on the stereotypical application of presumed group or personal characteristics. The second component of the test applies considerations similar to those applied by the United States Supreme Court by requiring the purpose of legislation to be identified and the relevance of the categorisation for achieving the purpose examined. Legislation will be valid if the legislative purpose is valid and the legislative categorisation of groups of people is relevant for achieving the purpose. 
L’Heureux-Dube J in Miron, Egan and Thibaudeau, applied a slightly different test by requiring consideration of whether the (allegedly discriminatory) categorisation ‘is capable of either promoting or perpetuating the view that the individual adversely affected by this distinction is less capable, or less worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration’.  According to Iacobucci J:
[Making this distinction requires] consideration of both the group adversely affected by the distinction and the nature of the interest adversely affected by it. The interaction of the group’s social vulnerability, in light of the social and historical context, and the constitutional and societal significance of the interest will determine whether the impact of the distinction constitutes discrimination.
Thus, the mere fact of categorising groups of people for unequal treatment may be valid if relevant for achieving a valid legislative purpose. It is unlikely, however, that categorisation on the basis of race will be relevant for allowing a valid purpose, unless for ameliorating disadvantage.
Although s51(xxvi) is not couched in terms of equality, indeed it was originally designed to allow unequal treatment, it arguably requires a rational relationship between the special racial characteristics of a group and the special legislative measures regarding the group. The characteristics which defined a race in 1901 were identified for the purpose of ranking races according to a hierarchy of social and biological superiority. The assumption of racial hierarchy meant that, at the time of Federation at least, s51(xxvi) legislation could, and indeed should, treat races unequally in terms of their basic rights and entitlements. Hierarchy no longer serves as a basis for defining race. Instead, it is now taken that racial difference does not imply racial superiority or inferiority. Legislation which now seeks to treat a race as inferior would be suspect and probably unconstitutional. The definition that the challenged legislation provides of the ‘race’ and the assumptions underlying the definition would invite scrutiny. And the connection between the characteristics that are singled out as defining the group as a race and the measures applied in the legislation regarding the race would also require strict scrutiny. In this context the United States and Canadian equal protection jurisprudence would be informative, but not decisive, in determining validity.
The race power can only be understood if the contingent nature of the term ‘race’ is itself appreciated. This can only be done if the historical circumstances of the creation and amendment of the race power are considered. Doing so reveals that the term ‘race’ is volatile and has altered substantially since it was first included in the Constitution. This factor profoundly affects the meaning and operation of s51(xxvi). Consequently any test for the application of the provision must accommodate the provision’s volatile nature whilst maintaining fidelity to the text of s51(xxvi). Gaudron J’s test in the Hindmarsh Island Bridge Case comes closest to achieving these objectives.