The Electoral and Political Implications of Reserved Seats for Indigenous Australians

Abstract

Reserved seats for Indigenous Australians have been proposed as a means by which to stimulate the relatively low voting participation of Indigenous citizens and to afford them more effective representation in parliament. Yet, the idea of dedicated Indigenous seats is regularly rejected by committees of inquiry and there are significant obstacles to their establishment. One of these obstacles is the lack of agreement or understanding about how dedicated seats would work in practice. We aim to help clarify this issue by testing in a modelling experiment how reserved seats might affect the outcome of elections and the composition of the Federal Parliament. We then assess the relative merits of each model for ease of implementation and likely acceptability for both the public and political elites and find that there are viable models with good chances of finding acceptance.

Keywords::

• Indigenous citizens
• representation
• reserved seats

Notes

¹It ultimately rejected them due to the high level of opposition expressed during public consultation (Chesterman 2006, 272–73).

²Goot (2006) provided an authoritative and detailed analysis of Aboriginal enfranchisement and its consequences.

³Neville Bonner was a senator from 1971 to 1983, Aden Ridgeway was also a senator from 1998 to 2005, while Ken Wyatt became the first Indigenous member of the House of Representatives in 2010.

⁴For a complete list of Indigenous state and territory MPs since 1974, see Appendix 1 at URL:

< https://hss.adelaide.edu.au/historypolitics/pdfs/Appendix_G_Evans_L_Hill.pdf>.

⁵However, the question of whether several Aboriginal representatives could ever represent Australian Aboriginal peoples is contentious due to the fact that the Indigenous population is not monolithic and is characterised by intense localism (Maddison 2010, 672; Orr 2003, 37; Sutherland and Russell 1998, 15).

⁶Stevenson and Jarred (2001, 19–25) provided a detailed examination of experiments in dedicated seats in other settings, namely, the United States, Sweden, Norway, Canada and New Zealand.

⁷The magnitude of this increase ranges between 6 per cent (Banducci, Donovan and Karp 2004) and 10–12 per cent (Brace et al. 1995).

⁸Political efficacy has two components: ‘internal efficacy’, which denotes a person's subjective sense of political competence, and ‘external efficacy’, which refers to the belief that government is responsive to ‘attempted influence’ (Parenti 1974, 160). A heightened sense of political efficacy enhances the motivation to vote.

⁹‘National surveys have for some time indicated that more than twice as many respondents think “government help for Aborigines has gone too far” as think it has “not gone far enough”’ (Goot 2006, 560).

¹⁰For further discussion of the legal implications of reserved seats in Australia, see Stevenson and Jarred (2001, 13–15).

¹¹Frank Brennan (1995, 201) has mooted a model whereby Indigenous electors could ‘vote not only for their State and Territory senators but also for four indigenous senators, one of whom would be a Torres Strait Islander. Alternatively, they could opt to vote for either indigenous senators or the regular senators for their territory or State’.

¹²Raw figures used to calculate the likely results of the hypothetical elections were obtained from the official Australian Electoral Commission website (URL: < http://www.aec.gov.au>).

¹³We are aware that this may inflate the numbers of Aboriginal voters, as the lower life expectancy among Indigenous citizens suggests a yield of a higher than average percentage of minors.

¹⁴This is because terms of state senators are staggered (six at each election), while territory senators are chosen at every election. Hence Chesterman's proposals affect the states and territories in different ways.

¹⁵Polling booths used to calculate the Aboriginal vote at the 2004, 2007 and 2010 elections can be found in Appendix 3 at URL: < https://hss.adelaide.edu.au/historypolitics/pdfs/Appendix_G_Evans_L_Hill.pdf>.

¹⁶This projection is derived from the fact that the proportion of Maori on the Maori roll is 57 per cent and rising since the introduction of mixed-member proportional representation. Given that our system is partly proportional, this augers well for rates of enrolment in Australia. Since there is less suspicion around reserved seats among the Indigenous people of Australia than among Maori, more Indigenous Australians would choose to be on a separate roll.

¹⁷Results of this and other elections can be found at URL: < http://www.electionresults.org.nz/electionresults>.

¹⁸There have been ‘at least seven’ Indigenous parties since the 1960s, though none are registered today (Orr 2003, 37).

¹⁹It is impossible to predict what effect dedicated seats would have on traditional Aboriginal allegiances and it is possible that, given the New Zealand experience, support for Labor would decline over time if viable Indigenous parties emerged in an exclusively Aboriginal electorate (thanks to two anonymous referees for raising this point). By the same token, it is worth noting that the Maori Party's support for the National Party government at the 2011 election probably cost it votes. It lost one seat each to Labour and Mana respectively, and its three successful candidates were returned with reduced majorities. Labour's two incumbents increased their majorities, which was generally not the case for Labour sitting members in non-Maori seats.

²⁰For example, in the two elections in which Aden Ridgeway ran for the Senate as a Democrat, he received no votes at Wallaga Lakes Koori Village (a predominantly Aboriginal polling place). Nor did the House Democrat candidate. In 1998 all 34 Senate votes went to Labor; for the House it was 30 Labor, 2 Greens and 2 Informal. In 2004 the Senate figures were 40 Labor, 6 Greens, 1 Liberal/National Party (LNP), 1 Christian Democratic Party and 1 informal vote. In the House the figures were 49 Labor, 4 Green, 2 Progressive Labor, 2 LNP and 1 informal vote.

²¹The assumption that ‘Aborigines would vote for an Aboriginal candidate, regardless of party considerations if one stood against a non-Aboriginal candidate… has been proved wrong by two unsuccessful Aboriginal candidates, G. Djerrkurra in 1980 and B. Liddle in 1984’ (Loveday and Sanders 1984, 68).

²²For detailed figures and explanation of how we arrived at our conclusions for the Senate, see Appendix 4 at URL: < https://hss.adelaide.edu.au/historypolitics/pdfs/Appendix_G_Evans_L_Hill.pdf>.

²³For a table showing ‘Predicted Seat Tallies by Party under Model 1 as Compared to Actual Senate Results’, see Table 2, Appendix 5 at URL: < https://hss.adelaide.edu.au/historypolitics/pdfs/Appendix_G_Evans_L_Hill.pdf>.

²⁴In a normal half-Senate election, there will usually be a 3–3 left–right split in most states. There have only been three exceptions to this trend in the last 18 elections: Queensland in 2004 and Tasmania in 2007 and 2010 (Mackerras 2011).

²⁵In predominantly Aboriginal polling places, Labor received 69.13 per cent of the two-party preferred vote in 2004, and 77.97 per cent of the two-party preferred vote in 2007.

²⁶The term ‘co-extensive’ means that the boundaries used for the five single-member House of Representatives seats in Tasmania are the same as those used for the multi-member Tasmanian House of Assembly seats. The creation of an Indigenous House of Representatives seat in Tasmania at the expense of a general seat would cause an upheaval in state boundaries, unless they did the same thing. Note: seats in the ACT Legislative Assembly are not co-extensive with federal seats.

²⁷Labor won Bass easily in 2010, so the exclusion of Aboriginal voters would have made no difference.

Additional information

Notes on contributors

Lisa Hill

Glynn Evans is a PhD candidate in the School of History and Politics, University of Adelaide. Lisa Hill is a Professor of Politics at the University of Adelaide. The authors wish to thank the Australian Research Council whose general funding for DP0985074 made the completion of this article possible.