What do we know about public attitudes to a First Nations Voice?
Francis Markham and William Sanders
15.05.2021
The cases for and against enshrining a First Nations Voice in the Australian Constitution are often said to hinge on its chances of success at a referendum. For example, in late 2020, when asked why the Coalition had rejected calls to enshrine a First Nations Voice in the Constitution, the Minister for Indigenous Australians Ken Wyatt argued that:
If you fail on a question for constitutional referendum, it is never resurrected. We only have to look at 8 that have been successful against 42 attempts. I don’t want this to fail.
On the other hand, high levels of public support for a First Nations Voice are sometimes cited as evidence of the imperative to hold such a reference. Hence, public attitudes to a First Nations Voice to Parliament remain a crucial consideration in public debate.
In this post, we update our findings from 2020 of the public opinion surveys with the most recent data. Levels of support for a Voice amongst those willing to offer an opinion remain very high, but a third of voters are looking for more information. Leadership from political elites remains key. We find that a Coalition-led referendum proposal would almost certainly succeed.
What did we find last year?
Last year, we examined evidence about whether a constitutionally enshrined First Nations Voice to Parliament was ‘capable of winning acceptance in a referendum’. We brought together and synthesised the results of 12 publicly published opinion surveys, conducted between June 2017 and June 2020, in a Centre for Aboriginal Economic Policy (CAEPR) Working Paper.
Our findings can be summarised in four main points:
A large number of people are undecided about a constitutionally enshrined Voice to Parliament. While the exact proportion of undecided voters depended considerably on question wording, around 30% of voters reported uncertain opinions in the best designed surveys.
Political leadership matters. Some voters appeared to be following party leaders in forming their opinions. Coalition voters were being led by party leaders into a position of opposing a Voice during 2018 and 2019, after being initially supportive in 2017. Labor and Greens voters, on the other hand, were following their leaders and consolidated support for a Voice over time.
Regardless of this ‘follow the leader’ effect, around 75% of voters who ventured an opinion were in favor of a constitutionally enshrined Voice in surveys held between June 2018 and June 2020.
Determining attitudes in the less populous states is difficult from national surveys, particularly in Queensland, Western Australia and Tasmania. The rules for constitutional alteration make attitudes in these states crucial, because majorities are required in four out of the six Australian states if a referendum is to be passed.
From these findings, we concluded that if the Coalition leadership returned to the idea of a constitutionally enshrined First Nations Voice with a more positive frame, it would be likely to sway a considerable bloc of voters towards supporting a Voice. This would make the success of a referendum all but assured. We argued that it was less clear what the outcome would be if a future Labor government put constitutional enshrinement of a First Nations Voice to a referendum in the face of a Coalition Opposition being either neutral or actively opposed.
New data since November 2020
Two new opinion polls have been published after our analysis was completed in November 2020.
2020 Australian Reconciliation Barometer
In late 2020, the Reconciliation Australia published their 2020 Australian Reconciliation Barometer. They asked respondents:
Are the following actions important or not for Australia as a nation?
To establish a representative Indigenous Body, to share the views of Indigenous Australians regarding Indigenous affairs and policies
To protect an Indigenous Body within the Constitution, so it can’t be removed by any Government
2021 Australian Constitutional Values Survey
The 2021 Australian Constitutional Values Survey by CQUniversity and Griffith University was conducted in February 2021 with a more detailed set of questions on the enshrinement of a First Nations Voice to Parliament. After being told that “The ‘Voice’ would be a representative Indigenous body that advises Parliament and the government when it makes laws, policies and programs that affect Indigenous people”, respondents were then asked: “At a referendum, would you vote in favour or against changing the Constitution to set up this advisory Indigenous ‘Voice’?” (A detailed discussion by Jacob Deem of this more nuanced survey has been published on the IndigConLaw blog.)
Findings
Support among committed voters for a constitutionally enshrined Voice remains remarkably high. More than 70% of committed voters have supported a constitutionally enshrined Voice to Parliament in all polls since June 2018. In the latest two surveys, support ranged from 70–80% of committed respondents. Heading into a referendum campaign, this will provide a significant baseline level of support for a Voice among the public (see Fig. 1).
Figure 1: Levels of support for a First Nations Voice to Parliament among 'committed' voters, 2017–2021
But much of the public is yet to make up their mind on the issue. Indeed, a question asked by the 2021 Australian Constitutional Values Survey revealed that three-and-a-half years on from the Uluru Statement, just 48% of the adult population had heard about this idea of an First Nations ‘Voice’ to Parliament. In the earlier surveys, around 30% of voters indicated that they ‘don’t know’ how they would vote in a referendum on the constitutional enshrinement of a First Nations Voice to Parliament and this remained the case in the 2021 Australian Constitutional Values Survey (see Fig 2).
This level of ‘undecided’ creates a considerable level of uncertainty. It suggests that there is significant scope for political leaders, both First Nations and non-Indigenous, to sway the public on the issue. In the context of a referendum, undecided voters are likely to be swayed as much by leaders they respect as by campaign materials and activities.
Figure 2: Percentage of the Australian population with an uncommitted position on a First Nations Voice to Parliament, 2017–2021
The recent Australian Constitutional Values Survey (ACVS 2021) also suggests that the public may be less polarized by political party than previous polling had implied. The ACVS 2021 finds that among those intending to vote for the Coalition who have made up their mind on the issue, 57% support a constitutionally enshrined Indigenous Voice to Parliament. This is in contrast to the 2019 ABC Vote Compass, which suggested that only 47% of Coalition voters with a position on the issue supported a referendum on Voice. Conversely, the 2019 ABC Vote Compass suggested that 87% of intending Labor voters were supportive of a Voice, but in the 2021 ACVS this figure was reduced to 81%.
We suspect differences in methodology contribute to this changed finding. The ABC Vote Compass is a heavily weighted ‘opt-in’ poll held in the context of a federal election. The ACVS 2021 may, methodologically, be a better indicator of underlying community sentiment.
Methodological issues aside, the published polling is consistent with a firming of support for a First Nations Voice among Labor and Greens voters, and significant levels of support among Coalition voters (See Fig. 3).
Figure 3: Levels of support for a First Nations Voice to Parliament among 'committed' voters, by federal voting intention, 2017–2021
ACVS 2021 is only the third published poll to give levels of support in the six Australian states. For a referendum to succeed, it requires majorities to vote ‘yes’ in at least four of the six states.
In recent decades, Victoria and New South Wales have been the most supportive states on constitutional change. Polling suggests that support for a First Nations Voice to Parliament follows this historical pattern, and is strongest in these two jurisdictions.
However, because Victoria and New South Wales dominate national surveys, little is known about levels of support in the remaining four states. Discerning opinion is a particular problem in Tasmania, as its small population makes it almost invisible in nationally representative polling samples. In the 2021 Australian Constitutional Values Survey, just 38 voters from Tasmania were sampled. This showed considerably higher levels of support than in the 2017 ACVS (see Fig 4). But the extremely large range of the error bars for these two Tasmanian findings must be noted. Error bars for South Australian findings in Fig 4 also have a large range, but levels of support do seem consistent and quite high in the three surveys of the last four years (see Fig 4 again).
In Queensland and Western Australia, support for Indigenous issues in general is often lower. For example, residents of these jurisdictions are 1.3 – 1.6 times more likely than residents of New South Wales to believe that ‘land rights have gone too far’ when asked in opinion surveys like the Australian Electoral Study. The same appears to hold with Voice. Both ACVS surveys in 2017 and 2021 find support among committed voters in these jurisdictions around 60-70%, but the other survey that allowed this finding in 2017 saw Queensland in the 50-60% range and Western Australia falling just below a majority (see Fig 4 yet again).
Figure 4: Levels of support for a First Nations Voice to Parliament among 'committed' voters, by state, 2017–2021
Conclusions
Six months and two surveys on from previous analysis, relatively little has changed in public opinion on a First Nations Voice to Parliament. Levels of support remain high among those willing to state their preference. However, levels of indecision remain high (and knowledge low), with around a third of voters unwilling to offer an opinion.
Support for a constitutionally enshrined First Nations Voice to Parliament remains patterned strongly along party lines. As such, a Coalition-led Referendum on the matter would almost certainly succeed. That is to say, a First Nations Voice to Parliament appears very capable of winning acceptance in a referendum if backed by the Coalition. The fate of a Labor-led referendum would be less clear, depending in part on how the Coalition positioned itself as neutral or opposed.
One blind spot remaining in our knowledge is levels of supports in the less populous states. A referendum is both a national vote, and six separate votes in six states, in which majorities much be secured in at least four states. We still lack sufficient information about levels of support outside the most populous states of New South Wales and Victoria. Future opinion research could usefully focus on determining attitudes in these four states — a majority in two states of the four less populous states would be enough to secure a Referendum win, but majority in three or even four would be highly desirable. Further opinion research on the Voice proposal would do well just to focus on the four less populous states, while assuming support in New South Wales and Victoria.
__
Francis Markham is a Research Fellow at the Centre for Aboriginal Economic Policy Research (CAEPR) at the Australian National University.
William Sanders is an Honorary Senior Fellow at the Centre for Aboriginal Economic Policy Research (CAEPR) at the Australian National University.
