Chesterman, John --- "Chosen by the People'? How Federal Parliamentary Seats Might be Reserved for Indigenous Australians Without Changing the Constitution" [2006] FedLawRw 9; (2006) 34(2) Federal Law Review 261

One notable outcome of the 2004 federal election is that no Indigenous Australian was elected to federal Parliament. Since July 2005, when Aden Ridgeway's term as Senator expired, there have been no Indigenous Australians among the 226 members of the national Parliament. This article explores briefly why this is the case, and then moves to consider the viability of Australia adopting a quota system whereby 'dedicated' or 'reserved' seats of Parliament would be created for Indigenous Australians.

	Home \| Databases \| WorldLII \| Search \| Feedback Federal Law Review

Even though lobbying on it dates back over 70 years, relatively little scholarly attention has been paid in Australia to this issue. The one significant exception to this is an article in 2001 by Alexander Reilly, which articulates in detail why the call for reserved seats for Indigenous Australians 'must be taken seriously as a matter of democratic theory'.^[2] In addition to Reilly's article, two brief articles in 1997 canvassed the possibilities and limitations of the idea,^[3]

and there have been several other mentions of it in academic journals, including by experts on dedicated seats in other jurisdictions, most notably, Catherine Iorns Magallanes.^[4]

The issue has also been discussed in reports at State and federal level that I shall consider shortly.

This article will briefly rehearse the philosophical justifications for reserving Indigenous parliamentary seats, which have been well described by Reilly, but it will then pay particular attention to investigating the apparent constitutional constraints that have historically been raised over the past 70 years whenever the idea of set seats has received considered attention. Reilly and others have looked briefly at this issue,^[5] but here I will tread new ground in drawing on archival material to describe the reasons given by various federal governments in rejecting the idea. I will then evaluate the ongoing significance of these constitutional arguments, before suggesting six different models by which reserved seats might be established without changes being made to the Constitution.

INDIGENOUS PARLIAMENTARY REPRESENTATION

There have only ever been two Indigenous federal parliamentarians. Neville Bonner, the first, represented the Liberal Party in the Senate between 1971 and 1983,^[6]

and Aden Ridgeway represented the Australian Democrats in the Senate between 1999 and 2005. Recently Mal Brough, who is now the Indigenous Affairs Minister, revealed that he may have some Indigenous heritage, but he does not identify himself as an Indigenous Australian.^[7] There has never been an Indigenous member of the House of Representatives, and there has never been an Indigenous federal politician representing the Australian Labor Party.

The odds have never favoured Indigenous people being elected to federal Parliament simply on the vote of Indigenous Australians. At around 2.4 per cent of the Australian population,^[8]

and geographically spread out, Indigenous people are a negligible voting bloc who, even if they voted as one, would be unable to influence more than a couple of lower house electorates, and possibly the Senate result in the Northern Territory.^[9] Even the most inventive drawing of electoral boundaries, which allowed for significant discrepancies in electorate size, would not alter this situation. In order to be elected, an Indigenous candidate for federal Parliament would have to receive a majority of their support from non-Indigenous Australians.

The electoral systems in use to elect the houses of federal Parliament result in nearly all lower house members, and the vast majority of upper house members, being representatives of major political parties. Thus a significant part of the reason for the lack of Indigenous federal politicians stems from the reluctance of major parties to preselect Indigenous candidates in 'winnable' seats.

The preferential system used to elect members of the House of Representatives requires successful candidates to achieve a majority of the preferred vote, and ensures the domination of major parties. While independents and small party representatives can occasionally gain lower house seats, the House of Representatives is dominated by large parties. Currently the 150 members include only three Independents and no minor party representatives.^[10]

Thus the failure of the major parties to preselect Indigenous candidates must be seen as a prime cause of their complete historical absence from that chamber.

The Senate, which since 1949 has used proportional representation as its electoral method, gives greater scope for the election of people from minority backgrounds for four linked reasons. The use now of party list (above the line) voting enables people from minority backgrounds to be elected via a party vote, rather than by an individual vote, which diminishes the likelihood of direct voter discrimination on the grounds, for instance, of race. Equally importantly this also diminishes the likelihood of direct party discrimination, as party fears about electors not wishing to endorse a person from a minority group are also likely to cause an under-representation of minority group members. (Parties once feared, of course, that voters would not elect female candidates, and this itself — aside from direct voter choice on the issue — was a significant cause of the historical lack of female representatives in Australia.^[11]) Linked to this is the fact that multi-member electorates, like the Senate, do not require a particular person to be held out as the sole representative of the electorate, which again diminishes the likelihood of direct voter, or party, discrimination against any one of the potential representatives.

Finally, the Senate's proportional representation electoral method also improves the election chances of small parties, which often have structures and pre-selection processes more accessible to members of minority groups. A full Senate election, following a double dissolution, sees the quota for a successful candidate at 7.7 per cent, while for a regular half-Senate election the quota is 14.3 per cent. It is true that far less than 14.3 per cent of the first preference vote in a half-Senate election can result in a person being elected to the Senate. To take an extreme case, in 2004 the Family First Party, which polled 56 376 first preference votes in Victoria, or less than two per cent, still managed to gain a Senate seat in Victoria, owing to the way preferences favoured it. But at the same election, it is worth noting, the Greens in Victoria polled nearly 9 per cent of the first preference vote and yet failed to gain a Victorian Senate seat.^[12]

Indigenous candidates have stood for election to the Senate as representatives of minor parties. Aden Ridgeway has represented the Australian Democrats, and Andrea Mason polled strongly when she stood for a South Australian Senate seat at the 2004 election representing the Family First Party.^[13]

But the Senate is still dominated by the major parties, whose representatives constitute 67 out of 76 Senators since July 2005.^[14]

Again, these parties have failed to encourage Indigenous candidates to represent them. This is, no doubt, partly a feature of the socio-economic dislocation suffered by Indigenous Australians, which results in few Indigenous people willing to negotiate the often restrictive party structures and pre-selection processes in order to put themselves forward as candidates. Added to this, one ought to note the extreme demands placed on the small number of Indigenous people who do put themselves forward to occupy public roles. Such are the current demands on them that people in this category, who would be well suited to political careers — one can think here of people such as Noel Pearson, Lowitja O'Donoghue, Patrick Dodson and Mick Dodson, to name a few — have resisted moving their careers into Parliament.

An equally significant factor in the small number of Indigenous representatives of the major parties is the parties' fear that the Indigeneity of a candidate may sway voters against them (as they once feared a woman's gender would). It is worth noting that while the Australian Labor Party has adopted 'affirmative action' procedures in order to ensure that at least 40 per cent of its representatives from 2012 are women, a similar process has not occurred for Indigenous representation.^[15]

Thus these related factors — the small percentage and geographic dispersal of the population that is Indigenous, the electoral systems that favour largish parties, the reluctance of Indigenous people to put themselves forward as candidates, and the reluctance of major parties to preselect Indigenous candidates — have conspired against the election of Indigenous candidates to federal parliamentary seats.

RATIONALE FOR ADOPTING A QUOTA SYSTEM

There are two related rationales for calling for set Indigenous seats in federal Parliament. The first concerns the importance of having a national Parliament that reflects the make-up of the society it governs. Reilly refers to various incarnations of this rationale, one of which is to suggest the intrinsic value of ensuring that parliaments are home to the range of identities that constitute a society. Another variation posits that the need for 'minority voices' in parliaments is 'a fundamental question of justice'.^[16]

The second rationale concerns not just the personal characteristics of parliamentarians, but the broader issue of the failure of Indigenous Australians to have their views represented in Parliament. It is one thing for an Indigenous person to be elected to Parliament. It is another for that Indigenous person to be elected as a representative of Indigenous people. Even if a handful of Indigenous candidates had been elected at the 2004 federal election, the current constitution of upper and lower house electorates means that they would not have been elected, to any significant degree, by Indigenous voters. Most of their constituents and electors would necessarily be non-Indigenous.

It is impossible to know what level of support both Neville Bonner and Aden Ridgeway had from Indigenous electors, as such statistics are not maintained, but one can be certain that both owed their Senate seats primarily to the votes of non-Indigenous Australians. Indeed, in Bonner's case, it is reasonable to conclude that parties other than the Liberal Party probably received far more Indigenous support during his years as a Senator.^[17]

Issues that are important to Indigenous communities are not subject to the same democratic and electoral pressure as issues of broader concern. A debate, for instance, during an election campaign about the need to abolish what Noel Pearson calls 'passive welfare' amongst Indigenous communities,^[18] will be far more influenced by what non-Indigenous voters think than by the views of Indigenous voters. The fact that 2004 saw the Howard government announce the end of the fifteen-year experiment with a directly elected Indigenous policy-making organisation, in the form of the Aboriginal and Torres Strait Islander Commission, only makes this issue more significant.

It is highly unlikely that an avowedly pro-Indigenous party, led by an Indigenous person, would be elected primarily by non-Indigenous electors. To illustrate the difficulty that confronts such a party, one can examine the performance of the Your Voice group at the 2004 election. Richard Frankland, the founder of Your Voice, stood for the Senate in Victoria on a platform solely devoted to improving the position of Indigenous Australians. Yet, with a reasonable deal of media coverage, his group attracted only 7266 first preference votes, or 0.24 per cent.^[19]

The most frequent criticism of the 'set seats' idea is the argument that the creation of set Indigenous seats necessitates the creation of set seats for other marginalised groups on the grounds, for instance, of ethnic identity or sexual orientation. Such a practice, the argument runs, would lead to the 'balkanisation' of Parliament, where Parliament is seen as a collection of vested interests in which representatives make decisions for their own group rather than for the public good.^[20]

The response to this argument is the claim that Indigenous Australians, as the descendants of people colonised without acquiescence, are a uniquely identifiable group with unique historical claims and particularly pressing policy needs. Their entitlement to a special voice in national political affairs is not shared by any other group. This is a claim that also receives support from emerging international law concerning self-determination.^[21] All of this is not to argue that there may not be convincing reasons one day for creating set seats for other marginalised groups. The point here is simply that the argument for set Indigenous seats is powerful.

Another criticism of the idea is that it would be tokenistic, in that a small number of Indigenous candidates might be elected who would then be marginalised within Parliament. That criticism would be valid if the representatives were to have observer and not voting status, but as the current numbers in the Senate attest, with the Coalition holding a one-seat majority, one vote can make a significant difference.

A final criticism is that the creation of an optional Indigenous roll would possibly encourage electoral rorting, whereby parties would seek to increase enrolment on the Indigenous roll to their electoral advantage. The response to this would be to simply state that the electoral law's restrictions on offering inducements would need to be enforced,^[22] with the possible extension of provisions to cover not only the offering of financial inducements to vote in a particular way, but to enrol on a particular roll. Other than that, one might simply see a party's encouragement for people to enrol on the Indigenous or general roll simply to be a part of the political process.

While the call for set seats is directly addressed at making the political system more relevant for Indigenous people, it is also worth noting the positive benefits that such a practice may have for non-Indigenous people. By encouraging Indigenous candidates to put their ideas forward, to be debated and determined by Indigenous voters, there will be greater engagement with the political process by Indigenous people. In time this could even lead, as arguably it has done in New Zealand, to a greater involvement by Indigenous Australians not only in the development of Indigenous policy debates, but in mainstream political debates, something which most people of most political persuasions would see as a positive development.

QUOTA SYSTEMS IN OPERATION

Quota systems guaranteeing parliamentary representation for certain groups are of course not new. Some of the countries that have reserved parliamentary representation for ethnic minority groups include Fiji, Lebanon and Singapore.^[23] The state of Maine in the United States has also reserved 'non-voting' Indigenous parliamentary seats, a practice that dates back to the 19^th century.^[24]

The system that warrants closest consideration so far as Indigenous Australians are concerned is New Zealand's, given that country's proximity to Australia and its similar colonial history. New Zealand has reserved seats for Maori since 1867, and currently reserves for Maori representatives seven out of a minimum 120 parliamentary seats.^[25]

Under New Zealand's electoral system census information and the numbers of Maori who choose to enrol on the Maori roll are used to calculate the number and location of Maori seats. Maori electorates and general electorates are required to have similar populations.^[26]

Maori New Zealanders can thus choose whether to be on the Maori roll or the general roll, and at present there are seven 'Maori electorates', with 62 'general electorates'. Under the Mixed Member Proportional (MMP) representation system a further 51 seats are allotted according to the party vote of electors, to ensure that political parties are proportionally represented in the House of Representatives.^[27]

In addition to the seven specified Maori representatives, the New Zealand electoral system has proven quite felicitous at enabling other Maori candidates to be elected. Maori can stand as candidates not just in the specified Maori electorates, but in the general electorates and on the party lists. The fact that New Zealanders register two votes for the one house of Parliament election — an electorate vote and a party vote — means that identified Maori electors can vote not just for a Maori representative in their Maori electorate but they can also vote for a party that has other Maori candidates on its list. In addition, of course, Pakeha and Maori electors on the general roll can elect Maori representatives either as general electorate representatives or from the party lists.

At present, following the 2005 New Zealand election, there are 21 Maori members of Parliament, representing a figure of over 17 percent of national politicians.^[28]

and thus Maori can now be said to have surpassed the proportion of seats that their population alone suggests they should hold.

The 21 Maori members include seven representatives of identified Maori electorates. But in addition to them, there are 14 'list' members.^[30] Clearly the 'list' vote mechanism has been particularly helpful to the electoral chances of Maori, with parties willing to put Maori in winnable list positions.

In Australia the most comparable vote to a New Zealand 'list' vote is the Senate vote, especially since the introduction of 'above the line' Senate voting in 1984.^[32] Since that time Australian electors have been able simply to vote for a party and allow that party to allot preferences. It is no coincidence that Australia's two Indigenous members of Parliament have both been Senators and not members of the House of Representatives. As the representatives of multi-member electorates, Senators, unlike members of the lower house, are not the sole representatives of their particular electorates. In this way both Aden Ridgeway and Neville Bonner did not have to hold themselves out as the sole representatives of their electorates, instead sharing their jurisdictions (New South Wales and Queensland respectively) with other Senators. Had they stood for lower house seats, Ridgeway and Bonner would need to have convinced non-Indigenous members of their electorates that they were able to represent all members of the electorate, a claim that many may not have disputed but that their political opponents may well have.

The main difference, of course, between the Senate vote in Australia and the 'list' vote in New Zealand is the quota needed before a candidate is elected. In New Zealand, parties only need five per cent of the vote to ensure they have representation in the Parliament.^[33]

But, as I have already mentioned, that amounts to about one-third of the vote needed to ensure the election of a Senator in a standard half-Senate election.

Debate has continued in New Zealand about whether the reserved seats should be continued for Maori, particularly given their ability to win election through non-specified seats.^[34] When it came to ensuring representation for other ethnic groups and women, the review committee in 2001 was adamant that other legislative quotas should not be implemented. It found that 'the use of the list provided the best means of ensuring a balance of representation, and that it remained the responsibility of political parties to ensure such a balance was maintained through candidate selection procedures'.^[35]

HISTORICAL CONSIDERATION OF QUOTA SYSTEMS IN AUSTRALIA

The search for direct Indigenous representation in federal Parliament has a long history in Australia. Bain Attwood and Andrew Markus trace the call back at least to the 1920s, and they have documented King Burraga's 1933 call for Aboriginal federal parliamentary representation.^[36] They have also examined, in considerable detail, the petition that famous Aboriginal leader William Cooper sought in the 1930s to place before Kings George V and VI, which called for 'a member of Parliament in the person of our own Blood, or White man known to have studied our needs and to be in Sympathy with our Race to represent us in the Federal Parliament'.^[37] Cooper's petition received detailed governmental consideration, with the Solicitor-General concluding that representation of Aborigines in federal Parliament in the manner sought was not constitutionally viable. He advised that, while the territories power in section 122 of the Constitution would allow Aboriginal Australians living in the Northern Territory or the Australian Capital Territory (ACT) to elect their own member,

In giving this advice, the Solicitor-General relied heavily on two sections of the Constitution that were amended in 1967. Section 51(xxvi) disentitled the Parliament from making specific laws with regard to Aboriginal people, and section 127 barred Aboriginal people from being taken into account '[i]n reckoning the numbers of the people of the Commonwealth'.^[39] As Frank McGrath has written, section 127 existed 'to determine … the amount of contribution to be made per capita by the people of the [S]tates to Commonwealth expenditure, and … the amount of any return to the States of surplus revenue'.^[40] Others have argued that section 127 was later incorporated for the dual reasons of applying 'not only to the financial clauses but to the "numbers" on which each State's membership in the House of Representatives would be based'.^[41]

In 1949 Shadrach James, from the Aboriginal Progressive Association of Victoria, wrote to Prime Minister Chifley seeking representation for Aboriginal people in federal Parliament.^[43]

Again the advice, this time from Prime Minister Chifley, was that such a proposal was constitutionally impossible:

The same year Doug Nicholls, from the Australian Aborigines' League in Victoria, called on the Prime Minister to allow all currently enrolled Aboriginal people to elect their own House of Representatives member.^[45] In the same year the Commonwealth electoral legislation was amended so that all Indigenous Australians who could vote at State level, or had served in the army, could exercise a federal vote (Western Australia and the Northern Territory continued to deny the franchise to Indigenous Australians until 1962, as did Queensland until 1965).^[46]

Nicholls followed his letter up with another, in which he quoted a resolution that had been adopted 'at a representative gathering of Aborigines at Mooroopna, Victoria'. That resolution stated:

Kim Beazley senior, who would play an important role in removing racial criteria from Commonwealth franchise laws in 1962, considered Nicholls' request 'completely just'.^[48] But as he had told Shadrach James, Prime Minister Chifley again simply responded that 'there is no provision in the Constitution' for such a practice.^[49]

Twelve years later, in 1961, when the House of Representatives established a select committee to consider enfranchising all adult Indigenous Australians, the issue of set seats was briefly canvassed. Again the legal opinion was negative, though this time the main constitutional impediments were not sections 51(xxvi) and 127, the constitutional provisions that would be amended in 1967.^[50] Rather the obstacles were other sections that remain in the Constitution today. I shall shortly consider these provisions, and the legal opinions regarding their implications for reserving parliamentary seats.

The issue of direct Indigenous representation in Parliament continued to be raised through the 1980s to the present day by activists, and even on one occasion, in 1983, by the New South Wales Aboriginal Affairs Minister.^[51] To give just a few relatively recent examples where notable Indigenous organisations have made this call: in 1993 the Council for Aboriginal Reconciliation advocated constitutional amendment in this regard;^[52] in 1995 ATSIC called for 'reserved seats in Parliament for indigenous Australians at both the Commonwealth and State level';^[53]

in the same year, the Council for Aboriginal Reconciliation argued there was a need to develop 'an educational strategy on the possibility of separate indigenous seats, based on an indigenous electoral roll, in the House of Representatives and in the Senate';^[54] and in 2000 the Council for Aboriginal Reconciliation called on governments to improve the representation of Indigenous Australians in Parliament.^[55]

One final possibility that has been raised in recent years regarding Indigenous representation in federal Parliament has been the notion that the Chair of ATSIC should be granted direct representation in Parliament via the territories power (section 122).^[56] The abolition of ATSIC has put paid to that idea, but any ex officio 'election' to Parliament using the territories power would carry its own legitimacy problems, and would likely run foul in any case of sections 7 and 24 of the Constitution, which require representatives to be 'directly chosen by the people'.

At State level there has been some consideration of reserving seats for Indigenous Australians. In 1965 Queensland became the last jurisdiction to enfranchise all adult Indigenous Australians, and the Queensland government briefly considered allowing Aboriginal people and Torres Strait Islanders to each elect one member of Parliament. It chose not to do so.^[57]

Two States in recent times have shown sufficient concern about Indigenous disengagement from political processes to warrant the briefing of parliamentary committees that have considered the creation of set seats at State level. In 1998 the New South Wales Legislative Council Standing Committee on Social Issues published a report entitled Enhancing Aboriginal Political Representation: Inquiry into Dedicated Seats in the New South Wales Parliament. That committee examined many problems posed by the reserving of set seats that are specific to New South Wales, including the likely need for a referendum to be passed if one or more seats were to be reserved for Indigenous Australians in either house of Parliament. The committee decided that more background work needed to be done before support for the proposal of specific Indigenous seats in the New South Wales Parliament could be given.^[58]

In 2003 the Queensland Legislative Assembly's Legal, Constitutional and Administrative Review Committee published a report on the involvement of Indigenous people in democracy. While Queensland does not have the same constitutional limitations that confront New South Wales, the committee did not support the establishment of specific State parliamentary seats for Indigenous people due to 'the degree of opposition to dedicated seats expressed during public consultation', though it indicated it would revisit this issue if Indigenous involvement in the democratic system did not improve in the following decade.^[59]

CONSTITUTIONAL IMPEDIMENTS AND PARLIAMENTARY POSSIBILITIES

In legal terms, the simplest way to create set seats would be to alter the Constitution to this effect. One could then avoid the constitutional limitations that shall be detailed in this section. But politically, this would be the most difficult of paths. Only eight out of the 44 national referendums have been successful in Australia, and even proposals that carry broad popular support, such as the republican movement, have failed to see that support translated into successful referendums.^[60]

The unlikelihood of a successful referendum on the issue of reserved seats is only one reason why any attempt to create reserved seats would have to work within current constitutional parameters. The other reason is that there may come a day when reserved Indigenous seats are no longer seen to be appropriate, and any constitutional entrenchment of the idea itself would be difficult to change.^[61]

There is reason to be confident that existing constitutional parameters — which I shall discuss shortly — can successfully be negotiated in creating set Indigenous seats in federal Parliament. As a general point, it is worth noting that the Constitution leaves a substantial amount of discretion to Parliament in the creation and operation of electoral systems. A variation of the phrase 'until the Parliament otherwise provides' is repeated in the key electoral provisions in the Constitution, including in sections 7, 24 and 29, as a result of which the Commonwealth Parliament appears to possess substantial power over the type of electoral systems to be used in Australia. Moreover, whenever constitutional grounds are relied upon to challenge electoral laws and practices, the High Court has almost always deferred to Parliament's power in this regard. As Gerard Carney has written, in Australia there has been a 'virtual lack of successful constitutional challenges to the electoral regime. The recognition of an implied freedom of political communication in 1992 is the one glaring exception to this trend'.^[62]

In cases dealing with disproportionate electorate sizes, the creation of electoral offences concerning voting methods, and regarding the registration of political parties, the High Court has steadfastly refused to invalidate the Commonwealth Parliament's electoral laws, even when these laws contravene fundamental political principles such as 'one vote, one value' (McKinlay^[63] and McGinty^[64]), criminalise the promotion of alternative voting methods (Langer^[65]), or cordon off a particular voting method ('above the line' voting in the Senate) for the election only of political parties with a minimum number of members (Mulhollond^[66]). As Carney suggests, only in the field of the 'implied freedom of political communication' (ACTV^[67]) has the Court in recent times substantially rejected a parliamentary initiative in the field of Commonwealth electoral law.^[68]

The trust placed by the Constitution's founders in the ability of the Commonwealth Parliament to determine electoral procedures has enabled evolving understandings of democracy to be reflected in Australia's electoral laws. Indeed, as Chief Justice Gleeson pointed out in Mulholland, had the specific requirements of voting eligibility as they existed in 1901 been enshrined then in the Constitution 'our system might now be at odds with our notions of democracy'.^[71] The scope given in the Constitution to allow the meaning and practice of democracy to develop now, I would submit, permits the creation of set Indigenous seats.

But before that conclusion can be persuasively argued, there exist three constitutional provisions that must specifically be addressed.

As I discussed earlier, up until the 1960s the existence of section 51(xxvi) and section 127 in the Constitution were drawn upon as constitutional reasons why set seats could not be created. The 1967 referendum repealed section 127 and amended section 51 (xxvi), thereby enabling the federal Parliament to enact Indigenous-specific laws.^[72]

But even before the 1967 referendum was held, experts were pointing to the existence of another constitutional hurdle to the creation of set seats.

Section 29: Can House of Representatives electorates overlap one another?

When the House of Representatives established a Select Committee in 1961 to investigate the granting of the vote to all adult Indigenous Australians, that committee devoted a small amount of time to considering the possibility of creating set seats for Indigenous Australians in the lower house, but soon shelved the idea.^[73]

The prime constitutional hurdle here was neither of the provisions that would be amended in 1967, but rather section 29, which continues to state the following:

During its inquiry, the Select Committee asked the Attorney-General's Department whether it was constitutionally possible to establish 'an aboriginal electorate throughout the Commonwealth, regardless of State boundaries, on the model of the New Zealand electorates for Maoris'. The Solicitor-General, K H Bailey, in specifically contemplating the creation of a House of Representatives electorate, found section 29 to provide the stumbling block. Bailey's response to the question was that:

Here Bailey advised that 'though it would be constitutionally possible in the Northern Territory to establish electorates for aborigines, it would not be possible in the States'. Again, section 29 was the main problem. According to Bailey, even the amendment of section 51(xxvi), which by then had been widely canvassed and which would occur six years later, would not change things, since section 29 presented insurmountable problems. According to Bailey:

Geoffrey Sawer, another eminent lawyer, had agreed that while the territories power (section 122) enabled the Commonwealth Parliament free rein in determining electoral boundaries and criteria in the Northern Territory, the same was not so with regard to the States. He argued:

A question thus remains about whether, in Sawer's words, 'the assumption of purely geographical "divisions"', precludes the creation of reserved lower house seats. In coming to their opinions over 40 years ago, both Sawer and Bailey took section 29's prohibition on creating seats 'out of parts of different States' to the conclusion that only geographical divisions were envisaged.

The contrary opinion is that section 29, as with the other electoral provisions in the Constitution, leaves a great deal up to the Parliament, and that only its specific prohibitions ought to impact upon the creation of lower house electorates. Under section 29 Parliament is empowered to determine where the divisions are in the States, and to decide how many representatives each division will possess. The only restriction on Parliament here is that it is not able to create divisions 'out of parts of different States'. So the only sort of electorate specifically proscribed by section 29 is of the Albury-Wodonga kind, where parts of more than one state are included. Beyond that the Constitution is silent.

This robust reading is supported by the attention section 29 has received in the years since Bailey and Sawer gave their opinions. The suggestion of one scholar has been that, under section 29, 'not only can Parliament construct electoral divisions, but Parliament can also fix the weight and composition of these divisions'.^[77]

Support for this robust reading of section 29 also comes from the most significant High Court case on the constitutional requirements concerning House of Representatives electorates. In McKinlay, Chief Justice Barwick found that section 29 was one of the sections that left 'with the Parliament the decision of a basis of distribution of the States into electoral divisions'.^[78] Indeed, as he pointed out, section 29 'left it in the first place to each State to determine the boundaries of the electoral divisions within the States and the number of members to be chosen by any division'.^[79] As he pointed out, at the time of federation, some States had 'dual member districts' while Tasmania had one 'six member district and ... a four member district'.^[80]

The Constitution's founders were thus content with a substantial lack of uniformity in electorate size and level of representation. When one combines this fact with the apparently broad power delegated to Parliament by section 29, and when one recalls the High Court's general reluctance to impose restrictions on the Commonwealth Parliament's decisions concerning electoral law, it would be a brave constitutional scholar who would now argue that Bailey and Sawer's interpretation of section 29 should hold sway. Indeed it now appears more than a little bold to suggest that an implication exists in section 29, and should be observed, that electorates must be geographically based and cannot overlap with one another.

As I shall suggest shortly, the robust reading of section 29 that I have articulated would even possibly enable the Indigenous populations of two or three entire states to form one electorate, so long as the electorate was not created from 'parts of different States'.

Sections 7 and 24: would set seats contravene the requirement that representatives be 'directly chosen by the people', or that there be 'Senators for each State'?

There are two other provisions in the Constitution that must be negotiated by any proposal to create set seats, sections 7 and 24. Section 7 specifies that:

Both sections 7 and 24 require members of Parliament to be 'directly chosen by the people'. This requirement, however, has not been held constitutionally to require equal electorate sizes, or 'one vote one value', which is not surprising given that section 7 enshrined unequal electorate sizes when it gave equal representation in the Senate to the states regardless of their populations.^[81] This is important for the set seats proposal, as it will be constitutionally possible for Indigenous electorates to be significantly less populous than non-Indigenous ones.

Another possible implication of the phrase 'directly chosen by the people' might be that 'the people' cannot be divided along racial lines. On this argument 'the people' can only be divided into the sorts of groups envisaged elsewhere in the Constitution, namely geographical electorates for the purposes of electing the Senate and House of Representatives. The response to this would be that so long as the election of Indigenous people to set seats was not 'indirect' — for instance through the election of candidates to an electoral college — then the broad scope given by the High Court to Parliament over electoral matters would continue in this regard.

A further restriction to the creation of set upper house seats is the requirement in section 7 that the Senate be made up of 'senators for each State'. Despite such clear wording, Parliament has been held to have the power (under section 122) to create Senators for the territories,^[82] and one probably unconstitutional, and in any case politically tenuous, possibility might be to use the territories power to enable all Indigenous Australians to elect a number of Senators (perhaps four), who could be officially classified as Senators for the Northern Territory and the ACT.^[83] This process would allow, for instance, one or two ACT Senators to represent Indigenous people there and in Victoria, South Australia, New South Wales and Tasmania, while one or two Northern Territory Senators could represent Indigenous people there and in Queensland and Western Australia. In addition to the likelihood that such a usage of the territories power would be deemed unconstitutional, it is unlikely that so constitutionally 'tricky' a manoeuvre would receive broad political support.

The existence of section 7 prevents some quite interesting possibilities for Indigenous representation, such as the creation, for instance, of three Senate positions for Indigenous representatives which might be voted for by Indigenous Australians nationwide. Section 7 would also prevent one Indigenous Senator representing more than one State. Such proposals would clearly require a constitutional referendum in order to be implemented.

Another restriction posed by section 7 is the requirement that 'senators shall be chosen for a term of six years' (which does not apply to territorial Senators). So a rotational system (that I will canvass shortly) could not operate on a three-yearly cycle, and would need to be on a six-yearly one.

WHICH HOUSE?

Before suggesting some models that would give set federal parliamentary seats to Indigenous Australians, it is worth investigating how many seats would, on a purely population basis, need to be set aside. The simplest way to work this out is to say that, at 2.4 per cent of the population,^[84] Indigenous Australians should have 2.4 per cent of the 226 federal seats, which would amount to five or six representatives. On a strictly proportional basis one would expect Indigenous Australians to have 1.8 Senators and 3.6 members of the House of Representatives.

Such a small number of representatives would mean, of course, that it would be extremely difficult to allow traditional groupings of Indigenous people to be reflected in electorates, as regardless of the particular model that is adopted, electorates would need to be at least the size of the smallest State.

Equally the small number of representatives would raise the question about whether effective, rather than merely token, representation would be provided, though this is something that would not properly be assessable until the proposal had been in operation for some years.

Given the constitutional restraints noted above, one cannot simply create national electorates for either house. So a more nuanced approach must be devised in order to work within current constitutional parameters. Before suggesting six models that arguably do this, one needs to bear in mind the relative Indigenous populations of the various States and Territories, in order to enable not only the constitutionality of the proposals to be judged, but also the extent to which they enable some sense of proportionate representation of Indigenous people throughout the country.

According to the Australian Bureau of Statistics, following the 2001 census the Indigenous population in the States and Territories was as follows:

State/Territory	Indigenous Population	Percentage of Population
New South Wales	134,888	2.0 %
Victoria	27,846	0.6%
Queensland	125,910	3.4%
South Australia	25,544	1.7 %
Western Australia	65,931	3.4 %
Tasmania	17,384	3.7%
Northern Territory	56,875	28.5%
Australian Capital Territory	3,909	1.2 %
Other Territories	233	n/a
Australia	458,520	2.4%

Another crucial issue to be considered would be a person's eligibility to enrol on an Indigenous roll. I would suggest the New Zealand model would provide a good starting point, under which Indigenous electors, at a regular census time, can choose to be listed on an Indigenous roll, entitling them to vote for the reserved seats. I will not here discuss what the eligibility criteria for this might be in Australia, but the criteria used until 2004 for ATSIC elections would clearly be an appropriate starting point for that discussion.^[86] In addition, it would be necessary to establish a minimum threshold enrolment figure in order for an election for an Indigenous seat to go ahead, but that could be worked out in due course. Further, electoral law would need to be tightened to prevent inducements being offered to encourage Indigenous people to enrol on the Indigenous roll, since optional enrolment (especially in numerically small electorates) carries justifiable concerns about rorting.

The Senate

The Senate was established with the dual role of providing 'revision and review in matters of general legislation' and to protect the interests of the States.^[87] While the party system has long overtaken any sense in which the Senate has operated to protect State interests, the chamber's 'review' function remains central to its justification and operation, even if there continues to be disagreement over what activities constitute a legitimate review role.^[88]

There is some sense, then, in which the Senate would be the appropriate chamber for direct Indigenous representation. While their numbers would never be sufficient to form a major voting bloc in the house of government, Indigenous Senators would be able to serve an effective 'review' role in the Senate, where their support could be crucial to the passage of legislation and where they could exercise other review functions in committee work and in questioning members of government.

The following models take into account the constitutional limitations mentioned earlier.

Model 1: An Indigenous Senator for each State

One possibility would be to have one Indigenous Senator representing each State. Indigenous electors in the Northern Territory could vote with South Australian electors (on the understanding that they should join with a bordering State, and of those States that border the Northern Territory, South Australia has the smallest Indigenous population), while Indigenous people in the ACT could take part in the New South Wales election (on the understanding that this would be constitutional when sections 7 and 122 are read together). The central problem with the proposal that each jurisdiction elect an Indigenous Senator is that some jurisdictions, for instance, Tasmania, would be electing an Indigenous Senator with a very small vote indeed. Even if a large part of the Indigenous population there nominated for the Indigenous electoral roll, the vote recorded would still be very small.

Moreover, if one assumes that the standard half-Senate elections are held, Indigenous electors would only be voting on the Indigenous roll at every second election, at which one of the six State Senators would be elected by as little, in some jurisdictions such as Victoria, as less than one per cent of electors.

Model 2: An Indigenous Senator for New South Wales, Western Australia, Queensland and the Northern Territory

Another possibility is to restrict the direct election of Indigenous Senators to those jurisdictions with high Indigenous populations, and to have one Indigenous Senator for New South Wales, Western Australia, Queensland and the Northern Territory. The obvious problem here would be that some Indigenous people would be unrepresented. Moreover, the proposal would have some unintended flow-on effects. For instance, if one Senate place were reserved for Indigenous people in New South Wales, then non-Indigenous people in New South Wales would be electing only five Senators at every second half-Senate election. Non-Indigenous Tasmanians, on the other hand, would be electing six Senators at every half-Senate election. As things stand at the moment, a Tasmanian's Senate vote is worth many times that of a New South Wales elector (given the fact that vastly different populations each elect the same number of Senators) and this proposal will only increase that differential for non-Indigenous voters.

Model 3: Three Indigenous Senators, elected on a rotating basis

One complex possibility would be to have three Indigenous Senators representing, on a rotational basis, the six States (with Territory electors joining with State electors as discussed in Model 1). Under this proposal a ballot would be drawn to determine a ranking among the States, and gradually each State would have its chance to elect one Indigenous Senator, who would serve for six years. Presuming a normal flow of Senate half-elections, three jurisdictions would elect Indigenous representatives at one election, with the other three each electing an Indigenous Senator six years later. For instance, this model might work as follows:

1^st half-Senate election: Queensland and South Australia (including Northern Territory) and New South Wales (including ACT) would each elect one Indigenous Senator

3^rd half-Senate election: Tasmania, Western Australia and Victoria would each elect one Indigenous Senator

In the event of a double dissolution election, whereby all Senate positions would be vacant, the next three jurisdictions due for an Indigenous Senator would simply elect Indigenous Senators. The process, under section 13, that follows a full Senate election is that the new Senators must be divided into long and short-term ones, so that the process of half-Senate elections can continue. The suggestion here would be to define all the Indigenous Senators as 'short-term' ones, on the understanding that this would increase the speed of the rotation system and allow jurisdictions more quickly to elect Indigenous representatives.

The main disadvantages of this system are its complexity, and the fact that, in the normal course of events (that is, in the absence of double dissolution elections), a jurisdiction only gets to elect one Indigenous representative once every four half-Senate elections, or every 12 or so years (though the jurisdiction's representative would be in place for half of that time). The problems identified in Model 2 are also relevant to some extent, in that at some elections, Tasmanian non-Indigenous electors might be electing six Senators, while New South Wales non-Indigenous electors might only be electing five. This, again, would mean even larger discrepancies than currently exist regarding voting power.

The House of Representatives

The other possibility is to provide set seats in the House of Representatives. This chamber was created to allow numerically congruent electorates to be each represented by one member, and, as the chamber of government, to allow the determination and expression of national will 'irrespective of State boundaries'.^[89]

There are two grounds on which one might argue that this chamber constitutes the appropriate one for direct Indigenous representation. First, many Indigenous communities face challenges that are not uniform. The challenges facing Indigenous people in remote areas are not the same as those facing people in urban areas. One line of argument would suggest that this would be the appropriate house in which to have reserved Indigenous seats, since members of the House of Representatives, in theory at least, represent specific communities which often have quite divergent interests (in a way that the States, as Senate electorates, do not).

Secondly, the desire to see Indigenous wellbeing as a national, rather than State, challenge again ties in with the original intention in establishing the House of Representatives.

Neither of these reasons, though, is particularly convincing. As the Senate does not operate as a States' house, Members of the House of Representatives do not really strive to represent the national interest anymore than do Senators. Moreover, the number of Indigenous representatives would have to be very small, meaning that their electorates would be large. This would disenable them, for instance, from having purely urban, or purely remote, constituents, and would in effect see them having at least State-wide electorates.

Rather than for theoretical reasons, it is for pragmatic reasons that the House of Representatives is the preferred place for direct Indigenous representation. The existence of section 7 of the Constitution in effect means that Indigenous Australians would either be over-represented (if each State had an Indigenous Senator), or that some Indigenous people would not be represented (if their State was too small numerically to justify an Indigenous Senator). For this reason alone, the lower house is to be preferred. The creation of set lower house seats was directly canvassed by the Select Committee in 1961, and this idea now bears renewed consideration.

The Australian Electoral Commission (AEC) works out a State's entitlement to lower house seats according to a formula that takes into account the 'nexus' provisions with the Senate (under section 24 of the Constitution the number of House of Representatives seats is as near as possible to 'twice the number of the senators' — a figure that does not include the Senators from the Territories). The combined population of the States (not Territories) is divided by 'twice the number of the [State] senators', and the result is a quota figure that is used to determine the number of lower house seats a State possesses. In 2003 the AEC determined the electoral quota to be 133 369. That is, the States (except Tasmania, which is guaranteed five seats by section 24 of the Constitution) received a House of Representatives member for every 133 369 people in their resident population.^[90]

It is important to note that the quota is worked out on total population figures, not adult figures. Thus the fact that the average age of the Indigenous population is younger than the non-Indigenous average would not be relevant to the determination of how many seats should be reserved for Indigenous Australians.

If one relies on the 2001 Indigenous population figure of 458 520, and divides it by the 2003 quota figure of 133 369, then Indigenous people would be entitled to three or four seats out of the 150 in the lower house.

Ideally one could create three or four electorates nationwide for the Indigenous population, and ignore State boundaries in doing so. But the key constitutional constraint here, as discussed earlier, is that House of Representatives seats cannot, under section 29 of the Constitution, 'be formed out of parts of different States'. That means that one cannot, for instance, join up parts of Victoria and South Australia to create Indigenous lower house electorates. This is significant, given that the Indigenous population in these States is much lower than the current figure of 133 369 used in 2003 to work out the number of lower house seats each State is to have. That probably does not, however, prohibit joining up Territories with States, as the next two proposals below do.

For the sake of argument, I am here presuming that a robust reading of section 29 (contrary to the opinions of Sawer and Bailey) would prevail, according to which the section, by prohibiting the creation of seats out of 'parts of different States', does not imply purely geographical (or non-overlapping) electorates. Thus Models 4 to 6 proceed on the assumption that section 29 does not prohibit the creation of State-wide lower house seats for Indigenous Australians in various jurisdictions in addition to the non-overlapping seats that currently exist.

Model 4: Six Indigenous House of Representatives seats

Under this model each State would have one Indigenous member of the House of Representatives, with Indigenous electors in the Northern Territory joining with South Australians to elect one representative, while ACT voters would join with New South Wales voters to elect one representative. In each case the entire jurisdiction, usually a State, would constitute the electorate. While electorates may vary significantly in size (from a total Indigenous population of 17 384 in Tasmania to 134 888 in New South Wales) this would, as discussed earlier, probably not present unsurmountable constitutional difficulties in terms of section 24. While current electoral legislation does require comparability in electorate size,^[91] an exception would need to be made here for the Indigenous electorates. On this point it is worth noting that Tasmania is guaranteed five lower house seats by section 24 of the Constitution, even though population figures suggest it should only have three or four.^[92]

The disparity in electorate sizes, and the fact that small electorates facilitate manipulation,^[93] constitute the main disadvantages of this system. This model, however, still ranks as the second best of the six models proposed, and would become the model of choice were Model 6 deemed to be unconstitutional. The reason for this is that at least Model 4 would allow all Indigenous Australians to be represented by an Indigenous member of Parliament. Moreover, in having six out of 150 lower house members, Indigenous Australians would be represented at a rate more proximate to their population figures than if they were represented by six out of 76 Senators (under Model 1).

Model 5: Four Indigenous House of Representatives seats: New South Wales, Queensland, South Australia and Western Australia

Another possibility is to allot lower house seats only to those States with large Indigenous populations. New South Wales (including the ACT), Queensland, South Australia (including the Northern Territory) and Western Australia might all be able each to elect one Indigenous lower house member, again with each jurisdiction constituting one electorate. While this lessens the likelihood of vastly disproportionate electorate sizes, it of course means that Indigenous people in some jurisdictions would be unrepresented.

Model 6: Four Indigenous House of Representatives seats: New South Wales, Queensland, North-Western and South-Eastern

A final possibility, and the preferred of the six models canvassed, would be to assume a very robust reading of section 29, and create four Indigenous lower house seats whereby all Indigenous electors could vote for a representative. The way to do this would be to argue that section 29, in requiring that lower house seats 'not be formed out of parts of different States', does not preclude entire States joining together to create one electorate. The argument here would be that the seats so created are not formed out of 'parts' of different States, but out of the entirety of different States. A High Court ruling would probably be sought on this issue, which would also give consideration to the phrase in section 24 that refers to the 'number of members chosen in the several States', which possibly suggests the founders of the Constitution only had in mind lower house seats that were fully contained within State boundaries.

If this model were deemed to be constitutional — and this might be stretching the High Court's deference to Parliament on electoral matters too far — then Indigenous electors could elect one lower house member in each of New South Wales (including the ACT) and Queensland, with the creation also of a North-Western seat (Western Australia and the Northern Territory) and a South-Eastern one (Tasmania, South Australia and Victoria).

While the actual size of electorates would not be known until Indigenous electors decided whether or not to enrol on the Indigenous roll, the electorate sizes under this model would be far more proportionate than under Model 4. The South-Eastern electorate would be smaller than average, but the fact that that seat would contain Indigenous electors from three States would justify its smaller size. The only disadvantage in this proposal is its constitutional uncertainty.

CONCLUSION

A political system has legitimacy problems when the most historically significant and perennially most marginalised and disadvantaged minority group is unable to have a single representative in the federal Parliament. As has been argued now for well over half a century, a special case exists for Indigenous Australians to have proportionate representation in the national governing body. While some of the constitutional reasons given over past decades to reject the set seats proposal do provide some ongoing dilemmas regarding just how the proposal might be implemented, these problems are not insurmountable and should be revisited. Six models have been canvassed that provide for Indigenous representation in federal Parliament. The preferred model would be Model 6, though that is also the most constitutionally tenuous one. Model 4, which exists on firmer constitutional grounds, would become the model of choice if Model 6 were deemed unconstitutional. Both models would enable all Indigenous Australians to elect Indigenous lower house members to represent them. Assuming their constitutionality, none of the models requires anything more than a simple piece of legislation in order to be implemented (and nor do they require any suspension or amendment of the Racial Discrimination Act). As well as making the proposal a politically viable one, the absence of any need for constitutional reform also ensures that the set seats proposal could be revised, or abolished, in future years if that were thought appropriate.

^[*] John Chesterman, Department of Political Science, University of Melbourne. I would like to thank the anonymous referees of this journal, in addition to Brian Costar and Greg Gardiner, for providing very helpful comments on a draft of this article, and I also thank Catherine Joyce, whose interpretation of s 29 of the Constitution was integral in developing some of the models discussed. The general argument canvassed in this paper was first broadly put by me in a newspaper article prior to the 2004 federal election: ‘Give Aborigines Their Own Seats’, The Age (Melbourne), 25 September 2004.

^[7] 'Minister Identifies with New Portfolio', Weekend Australian (Sydney), 28-29 January 2006, 1.

^[9] It is worth noting that the extent of their geographic dispersal would not always have led to Indigenous people as a bloc having such minimal electoral impact, had the Australian Constitution and electoral laws not been specifically racist. Brian Costar has noted that, at the turn of the 20^th century, if Indigenous people had been enfranchised at Commonwealth level, and if section 127 of the Constitution had required Indigenous people to be included in population statistics, then Western Australia, Queensland and New South Wales would probably have been allotted extra House of Representatives seats (with a corresponding loss of seats from South Australia and Victoria). Moreover, he argues, for a short time 'it is quite possible that there would have been a de facto … Aboriginal seat in each of Queensland and Western Australia'. See Brian Costar, '"Odious and Outmoded"?: Race and Section 25 of the Constitution', in John Chesterman and David Philips (eds), Selective Democracy: Race, Gender and the Australian Vote (2003) 93–4.

^[10] This figure includes National and Country Liberal Party members as members of the Coalition. See House of Representatives, List of Members by Political Party <http://www.aph.gov.au/house/members/mi-party.asp> at 12 April 2006.

^[11] See Deborah Brennan, 'Women and Political Representation', in John Summers, Dennis Woodward and Andrew Parkin (eds), Government, Politics, Power and Policy in Australia (7^th ed, 2002) 277–93.

^[17] See Scott Bennett, Aborigines and Political Power (1989) 128. See also ibid 85.

^[19] For the platform of this group, and an example of the media coverage given to it, see Graham Matthews, 'Indigenous Australians need a voice', Green Left Weekly, (online edition, 22 September 2004) <http://www.greenleft.org.au/back/2004/599/599p13.htm> at 13 October 2004. For the election result, see Australian Electoral Commission, above n 12 .

^[20] Reilly, above n 2, 87–9, 96. It is worth noting that one current proposal exists for the creation of a Senate seat for Australians living overseas. See Andrew Leigh, New Voting Rights for the Australian Diaspora, Australian National University, Democratic Audit of Australia (2004) <http://democratic.audit.anu.edu.au/papers/200407_leigh_expat_rep.pdf> at 28 July 2006.

^[21] For instance, art 19 of the Draft Declaration on the Rights of Indigenous Peoples (1994) E/CN.4/SUB.2/1994/2/ADD.1 holds that 'Indigenous peoples have the right to participate fully, if they so choose, at all levels of decision-making in matters which may affect their rights, lives and destinies through representatives chosen by themselves' (emphasis added). See also Elliott Johnston, Royal Commission into Aboriginal Deaths in Custody, National Report (1991), vol 2, chap 20; Reilly, above n 2, 91; Iorns [Magallanes], above n 4, [15].

^[23] Legislative Council Standing Committee on Social Issues, Parliament of New South Wales, Enhancing Aboriginal Political Representation: Inquiry into Dedicated Seats in the New South Wales Parliament (1998) 21.

^[24] See S Glenn Starbird Jnr, Donald Soctomah and Sue Wright, A Brief History of Indian Legislative Representatives, Maine State Law and Legislative Reference Library <http://www.maine.gov/legis/lawlib/indianreps.htm> at 26 November 2004. See also ibid; Iorns [Magallanes], above n 4, [34]–[38].

^[25] New Zealand Parliament, Electorate Profiles 2005: Maori Electorates <http://www.ps.parliament.govt.nz/electorate-maori.htm> at 10 April 2006. See also Iorns [Magallanes], above n 4, [17]ff. Since writing this article I have been pointed to an engaging and insightful draft article on New Zealand's Maori seats by Andrew Geddis, entitled 'A Dual Track Democracy? The Symbolic Role of the Maori Seats in New Zealand's Electoral System'. The draft article (on which the author welcomes comments) is available at <http://www.law.ucla.edu/docs/a_dual_track_democracy.doc> at 11 April 2006.

^[27] See Elections New Zealand, Reviewing Electorate Numbers and Boundaries <http://www.elections.org.nz/electorates/reviewing_electorates.html> at 10 April 2006. The number of seats in Parliament can exceed 120, where a party wins more electorate seats than its party vote suggests it should.

^[31] MMP Review Committee, above n 26, 46. This is quoted also in Legal, Constitutional and Administrative Review Committee, Queensland Legislative Assembly, Hands on Parliament: A Parliamentary Committee Inquiry into Aboriginal and Torres Strait Islander Peoples' Participation in Queensland's Democratic Processes (2003) 57 (see also 53).

^[36] Bain Attwood and Andrew Markus (eds), The Struggle for Aboriginal Rights: A Documentary History (1999) 16, 73.

^[37] A copy of Cooper's petition is contained in A431/1, 1949/1591, National Archives of Australia (hereafter NAA), Canberra. The petition is discussed in memorandum from J McEwen, Minister for the Interior, to Cabinet, 1 February 1938, A431/1, 1949/1591, NAA, Canberra. A slightly different form of the petition appears in Attwood and Markus, above n 36, 144. It is worth noting, as Attwood and Markus's documents show (for example at 88 and 115), that some prominent Indigenous people have opposed the idea. See also Bain Attwood, Rights for Aborigines (2003) Chapter 3; Bain Attwood and Andrew Markus (eds), Thinking Black: William Cooper and the Australian Aborigines' League (2004) 6–11, 35–6; Reilly, above n 2, 82.

^[38] Opinion of Solicitor-General George S Knowles, 14 January 1938, A431/1, 1949/1591, NAA, Canberra.

^[40] Frank McGrath, The Framers of the Australian Constitution: Their Intentions (2003) 197 (see 193–200 on this general point).

^[41] J A La Nauze, The Making of the Australian Constitution (1972) 68 (the note).

^[42] Letter from W Ferguson, Aborigines Progressive Association, to Governor-General of Australia, 30 August 1940, A431/1, 1949/1591, NAA, Canberra.

^[43] Letter from Shadrach James, Aboriginal Progressive Association of Victoria, to Prime Minister Chifley, 21 February 1949, A432, 1949/258, NAA. James is on record as favouring the idea as far back as 1930: see Attwood and Markus, above n 36 , 142.

^[44] Letter from Prime Minister Chifley to Shadrach James, 28 March (specific date unclear on document) 1949, A432, 1949/258, NAA.

^[45] Letter from Doug Nicholls, Australian Aborigines' League, to Prime Minister Chifley, R G Menzies and A W Fadden, 1 July 1949, A431/1, 1949/1591, NAA, Canberra. See also Attwood and Markus (eds), above n 36, 164–5; Reilly, above n 2, 82–3.

^[46] See John Chesterman and Brian Galligan, Citizens Without Rights: Aborigines and Australian Citizenship (1997), 157–62, 169–77.

^[47] Letter from Doug Nicholls, Australian Aborigines' League, to Prime Minister Chifley, 6 August 1949, A431/1, 1949/1591, NAA, Canberra.

^[48] Letter from Kim Beazley Snr to Prime Minister Chifley, 14 September 1949, A431/1, 1949/1591, NAA, Canberra.

^[49] Letter from Prime Minister Chifley to Doug Nicholls, date unclear (but in response to Nicholls' letter of 6 August 1949), A431/1, 1949/1591, NAA, Canberra.

^[50] Commonwealth House of Representatives, Report from the Select Committee on Voting Rights of Aborigines, part 1, Parl Paper (1961) vol 2, 8.

^[54] Council for Aboriginal Reconciliation, Going Forward: Social Justice for the First Australians, Recommendation 15. See also McGill, above n 3.

^[55] For example, Council for Aboriginal Reconciliation, Recognising Aboriginal and Torres Strait Islander Rights (2000) Chapter 3 <http://www.austlii.edu.au/au/other/IndigLRes/car/2000/9/pg6.htm> at 1 October 2004. See also Legal, Constitutional and Administrative Review Committee (Qld), above n 31, 53.

^[57] John Chesterman, Civil Rights: How Indigenous Australians Won Formal Equality (2005) 159–60.

^[58] Legislative Council Standing Committee on Social Issues (NSW), above n 23, especially Chapters 9 and 10. See also Iorns [Magallanes], above n 4 [46]–[52].

^[59] Legal, Constitutional and Administrative Review Committee (Qld), above n 31, 53, 56. The constitutional issues are well canvassed in the submission (no. 30) to the committee from the Gilbert and Tobin Centre of Public Law, University of New South Wales, 10 April 2003 (written by George Williams, Sean Brennan and Nicholas Hume) <http://www.gtcentre.unsw.edu.au/publications/docs/pubs/2003_queenslandIndigenousRep.doc> at 28 July 2006. See also Iorns [Magallanes], above n 4, [53]–[67].

^[60] See Ian McAllister, 'Elections Without Cues: The 1999 Australian Republic Referendum', (2001) 36 Australian Journal of Political Science 247.

^[62] Gerard Carney, 'The High Court and the Constitutionalism of Electoral Law', in Graeme Orr, Bryan Mercurio and George Williams (eds), Realising Democracy: Electoral Law in Australia (2003) 170, 171.

^[74] Opinion of K H Bailey, 7 June 1961, in ibid, appendix vi, 42 (original emphasis). See also ibid 8. My thanks here to Russell McGregor, who reminded me of the Select Committee's consideration of this issue.

^[75] Opinion of K H Bailey, 3 August 1961, in ibid, appendix v, 40–1 (original emphasis). See also ibid 8.

^[81] McKinlay [1975] HCA 53; (1975) 135 CLR 1. See submission (no. 30) from the Gilbert and Tobin Centre of Public Law, University of New South Wales, to the Legal, Constitutional and Administrative Review Committee (Qld), above n 59.

^[83] This suggestion was made to me during question time after I had presented a talk on this topic to the Rotary Club of Balwyn, 15 February 2005.

^[86] It is worth noting that these eligibility criteria have not been without their problems. For instance, there has been much debate regarding the eligibility of people to vote and stand in ATSIC elections in Tasmania in recent years. See Will Sanders, 'The Tasmanian Electoral Roll Trial in the 2002 ATSIC Elections' (2004) 63 Australian Journal of Public Administration 51.

^[87] John Quick and Robert Garran, The Annotated Constitution of the Australian Commonwealth (1976 [1901]) 418.

^[88] Stanley Bach, Platypus and Parliament: The Australian Senate in Theory and Practice (2003) 144, 156.

Federal Law Review

Chesterman, John --- "Chosen by the People'? How Federal Parliamentary Seats Might be Reserved for Indigenous Australians Without Changing the Constitution" [2006] FedLawRw 9; (2006) 34(2) Federal Law Review 261

'CHOSEN BY THE PEOPLE'? HOW FEDERAL PARLIAMENTARY SEATS MIGHT BE RESERVED FOR INDIGENOUS AUSTRALIANS WITHOUT CHANGING THE CONSTITUTION

INDIGENOUS PARLIAMENTARY REPRESENTATION

RATIONALE FOR ADOPTING A QUOTA SYSTEM

QUOTA SYSTEMS IN OPERATION

HISTORICAL CONSIDERATION OF QUOTA SYSTEMS IN AUSTRALIA

CONSTITUTIONAL IMPEDIMENTS AND PARLIAMENTARY POSSIBILITIES

Section 29: Can House of Representatives electorates overlap one another?

Sections 7 and 24: would set seats contravene the requirement that representatives be 'directly chosen by the people', or that there be 'Senators for each State'?

WHICH HOUSE?

The Senate

Model 1: An Indigenous Senator for each State

Model 2: An Indigenous Senator for New South Wales, Western Australia, Queensland and the Northern Territory

Model 3: Three Indigenous Senators, elected on a rotating basis

The House of Representatives

Model 4: Six Indigenous House of Representatives seats

Model 5: Four Indigenous House of Representatives seats: New South Wales, Queensland, South Australia and Western Australia

Model 6: Four Indigenous House of Representatives seats: New South Wales, Queensland, North-Western and South-Eastern

CONCLUSION

'CHOSEN BY THE PEOPLE'?
HOW FEDERAL PARLIAMENTARY SEATS MIGHT BE RESERVED FOR INDIGENOUS AUSTRALIANS WITHOUT CHANGING THE CONSTITUTION