Inside Story

Why did Australia reject the Voice?

Three books, two articles and a report offer a range of explanations

Tim Rowse Books 28 June 2024 5017 words

Winning slogan: a No to the Voice rally in Melbourne on 23 September last year. Alexander Bogatyrev/SOPA Images/Sipa USA


On 14 October last year the Yes campaign needed just under eight million votes distributed in a way that achieved a majority in four or more states. But only 6.35 million people voted Yes, and in no state were they in the majority.

The result was no surprise: pollsters had measured a steady decline in Yes support since October 2022, with roughly twenty percentage points shed in the twelve months leading up to the vote. But those who had been confident in 2022 were shocked by the speed of that decline. “Seldom does public opinion shift so dramatically in such a short period of time on questions of such importance,” says one analysis of the campaign.

Some interpretations of the defeat come to us as “lessons.” In their new books on the Voice, three authors — Andrew Gunstone, Frank Brennan and Damien Freeman — use the experience of October 2023 to propose better ways of handling the continuing politics of nation-building. Each author is more critical of the Yes campaigners than of the advocates of No.

Andrew Gunstone, associate deputy vice-chancellor for reconciliation at Federation University, has spent much of his career advocating and surveying Australians’ views about reconciliation. In Reflections on the Voice: During and After the Campaign, he reports that his surveys in 2005, 2010, 2015 and 2020 showed (in his summary) “how racist many non-Indigenous people are towards Indigenous issues as well as how ignorant they are and also how lacking they are in a sophisticated understanding of what equality is.” These findings didn’t however destroy Gunstone’s hope that voters would pass the referendum.

What made Gunstone hopeful was his long involvement in the propagation of Reconciliation Action Plans, or RAPs, among governments, businesses and community groups. By 2023, he reports, 2400 organisations had adopted RAPs (though later in his book — post-referendum — he writes that 5000 organisations have RAPs). Survey evidence suggests that staff in organisations with RAPs have “significantly higher levels of understanding and support for reconciliation.” Many organisations with RAPs pledged support for Yes in the October 2023 referendum. With “over four million people working or studying at a RAP organisation,” says Gunstone, the Yes campaign seemed to have a good chance.

What, in his view, went wrong? He makes two points.

The first is that many who support “reconciliation” lack a “sophisticated understanding” of equality. Australians, he argues, don’t want equality between Indigenous and non-Indigenous Australia to go beyond “equal rights and opportunities” (a widely supported value) to take in “special rights” such as native title, reserved seats in parliament or a treaty.

Gunstone’s surveys found that 97 per cent of respondents agreed with the statement, “As far as possible, all Australians should have equal rights and opportunities.” But only 45 per cent agreed that “Indigenous people should be entitled to special rights, such as native title,” only 42 per cent want the federal government to “negotiate a treaty with Indigenous people,” and only 32 per cent agreed that special rights should be in the form of “reserved seats in parliament.”

Had more Australians been “sophisticated” in their understanding of equality, Gunstone concludes, they would have supported a constitutionally entrenched channel of representation as a “special right.”

Survey data gathered by ANU’s Nick Biddle and Ian McAllister and published in their recent article “Safety or Change? The 2023 Australian Voice Referendum” indirectly supports Gunstone’s point. “Yes voters were more than twice as likely to be drawn from tertiary-educated voters as voters with no educational qualifications,” they write, and “the more attention a voter paid to the campaign the more likely they were to vote yes.” What Gunstone calls a “sophisticated” understanding of equality appears to be associated with formal education and engagement with the debate.

Second, Gunstone sees superficiality in the reconciliation movement: for many, he says, reconciliation is merely a desire for better relationships. Signing up to reconciliation has not required a person or organisation to commit to the “sophisticated” understanding of equality. Thus, a significant minority of organisations with RAPs, including universities, did not declare support for the Voice.

Reconciliation is popular partly because it isn’t demanding, says Gunstone. He sees limits in “‘performative reconciliation” — “organising morning teas for Reconciliation week, or wearing trendy T-shirts” — and he believes reconciliation has become a civic ideal without promoting “substantive reconciliation” — that is, support for special rights for Indigenous Australians.

Thus it was possible to convince many that the Voice, as something “special,” was a threat to reconciliation. Again the findings of Biddle and McAllister are pertinent:

While the goal of reconciliation is broadly supported across the population, the desire to ensure equality of treatment by the political institutions proved to be a stronger driver of the vote. For many voters, equality of treatment was seen as more important than giving special rights to one group, however deserving.

On this view, rather than extinguishing “reconciliation,” as some have said, the No majority affirmed a certain understanding of it.

Gunstone continues to believe in the political significance of the reconciliation movement while hoping to reform it from within. He urges Australians (including those who think they understand and support reconciliation) to further “educate themselves,” so that they grasp substantive reconciliation, including commitment to the Uluru Statement.

If, for Gunstone, the referendum demonstrated the superficiality of Australians’ progressive politics, for Jesuit priest Frank Brennan, a longstanding observer-participant in Indigenous politics, it manifested the political ineptitude of the Albanese government. In particular, he writes in his latest book, Lessons from Our Failure to Build a Constitutional Bridge in the 2023 Referendum, the government failed to cultivate bipartisanship.

The referendum was doomed to defeat, says Brennan, once the Liberal Party declared on 5 April 2023 that it would campaign for No. Believing that it was far from inevitable that Liberal leader Peter Dutton would declare his party opposed, Brennan holds the government responsible for missed opportunities to secure Coalition support.

Survey evidence does indeed show that Dutton’s declaration of opposition made a difference. Biddle and McAllister report that

by the time the referendum was held in October, just 8 per cent of Coalition voters supported the change compared to 62 per cent of Labor supporters. Indeed, the proportion of Coalition voters supporting the yes side declined by 30 percent over the four survey waves [January, April, August and October] compared to a decline of 16 per cent among Labor supporters.

According to Brennan, the government could have reached out to the opposition in either or both of two periods. After announcing the draft Section 129 at Garma in July 2022, the prime minister could have demonstrated his openness by initiating a formal process to consider criticisms and discuss alternative words. The debate that took place in civil society was no substitute for some kind of constitutional convention. The result of such a formal process would have been an amendment proposed by the Australian parliament, says Brennan, rather than carrying the burden of being explained and defended as a Labor idea.

In September, compounding that problem, the government appointed an Indigenous working party that knew public opinion polls to be favourable and was sceptical of the need for bipartisanship. Their view that the Coalition was beyond wooing was soon encouraged by the National Party’s decision at the end of November to oppose the Voice.

Nonetheless, Brennan suggests that bipartisanship remained possible in the second period, the summer of 2022–23, when the government was formulating the Constitution Alteration (Aboriginal and Torres Strait Islander Voice) Bill. During hearings on the bill in April and May before a joint parliamentary committee, legal experts gave much attention to what Brennan refers to as the “looseness” of the draft announced at Garma.

That “looseness” lay in the italicised words in this sentence: “The Aboriginal and Torres Strait Islander Voice may make representations to parliament and the executive government on matters relating to Aboriginal and Torres Strait islander peoples.” Like Brennan, some moderate Liberal MPs, while supporting the principle of a constitutional Voice, felt these words must be reconsidered if it were to be feasible as an instrument of government.

To put the phrase “executive government” in the constitution, Brennan argued, would entangle a “quarter of a million public servants” (“the executive government”) with a Voice that was entitled to “make representations” about all their administrative decisions. How would the Voice “be apprised of the matter when a public servant was making a decision,” asks Brennan, and would those public servants be constitutionally obliged to “entertain any representation then received from the Voice”? His questions conjured up the spectacle of thousands of public servants obliged to inform the Voice about administrative decisions they were soon to make and to which the Voice would then feel entitled (and even obliged?) to pay attention.

In Brennan’s view, this quagmire could be avoided in one of two ways. One was to omit the phrase “executive government” — effectively returning to the Referendum Council’s July 2017 recommendation that the Voice be “to Parliament” only. The other was to find words to restrict the scope of the phrase “executive government.” The Voice’s representations to the executive should be to “ministers of state” only, Brennan suggested, and his book documents his persistent advocacy of this change during 2023.

Although it heard the fears of Brennan and others, the parliamentary committee was assured by legal experts that the words tabled in parliament on 30 March, if litigated, would not be interpreted by the High Court in a way that made governing impractical. The committee’s Labor majority endorsed the constitutional amendment to which the government had been publicly committed since 23 March.

Attorney-general Mark Dreyfus sought to address the amendment’s contentious reference to “executive government” in his second reading speech by declaring that when parliament legislated the Voice it would “determine whether the executive government is under any obligation in relation to representations made by the Voice.” Dutton dismissed this assurance, and Brennan concurs with him: “Nothing said or done by parliament could impose limits on the scope of broad language in a proposed constitutional change.”

According to Brennan, the government’s stance in the parliamentary committee removed the last opportunity for moderate Liberals to bring a revised amendment to the Liberal party-room and persuade Dutton to drop his opposition to what he called the “Canberra Voice.” “From there,” writes Brennan, “the referendum was doomed.” But was it really the “executive government” question that sank the referendum?

A constitutional convention in 2022 or a consensus-seeking parliamentary inquiry in 2023 could certainly have found new words for the destination and scope of the Voice’s representations. But that doesn’t mean the Liberal Party would have endorsed the revised wording. In fact, Brennan himself gives reasons to question his thesis — reasons that raise doubts about his characterisation of the Coalition.

Crucially, Dutton had other objections. By January 2023 he had gained authority as a new Liberal leader by demanding that the government show voters a model of the Voice. A bipartisan approach by Albanese could have met Dutton’s demand for detail by committing to legislating the model presented in Tom Calma and Marcia Langton’s report on the Indigenous voice co-design process. But the government wanted the debate to be about the principle of the Voice not about the details of the subsequent legislation.

The Calma–Langton model contained no end of detail, but the Coalition would have been free to decide (in good faith or not) how to apportion its commentary between the principle of the Voice and its design. Brennan — having collaborated in the writing of Calma and Langton’s report — has no empathy for this dilemma. Remarking that the Yes campaign was handicapped “without a fleshed out model of a Voice for voters to understand,” Brennan has nothing critical to say about how the Coalition used its tactical advantage as the interrogator — virtuously aligned with the enquiring voter — persistently demanding “details” from a government promoting a principle.

Curiously, Brennan doesn’t pursue the implications of his remark that “the opposition parties had no interest in amendments which might render the proposal more acceptable to voters on the more conservative side of politics.” In fact, his remark is more consistent with what the Yes-supporting Liberal MP Bridget Archer told Lech Blaine: “Dutton opposed the Voice for purely political reasons… It’s straight out of the Tony Abbott playbook of just oppose everything.”

“Oppose everything” doesn’t quite capture the subtlety of Dutton’s position. Continuing the Morrison government’s stance, the Liberals said that they would legislate a Voice. Dutton was sufficiently pro-Voice to endorse a truncated version of the Calma–Langton model in which the formation of regional Voices would not be followed by the formation of a national Voice. What differentiated Liberal from Labor was not whether Australia would benefit from an Indigenous Voice but whether such a Voice should be in the constitution.

The official Yes pamphlet devoted only a few words to this central issue:

Putting the Voice in the Constitution gives it stability and independence, now and into the future. This means the Voice can give frank advice, without getting caught up in short-term politics. We can’t solve all the challenges Indigenous Australians face overnight. We need action now, as well as planning for the long term. That’s why we need a Voice that can’t be abolished with the stroke of a pen.

In a second reading speech of just under 2500 words, Dreyfus devoted even fewer words (eighteen) to the question of why only a constitutionally entrenched Voice was worth having: “votes in this parliament or the stroke of a minister’s pen have seen previous bodies abolished or defunded.”

To evoke the constitution as a brake on government caprice is to mobilise a certain understanding that document. But that is not the only way that Australians think about the constitution; they also see it as a document guaranteeing certain values. Both the Yes and the No campaigns evoked the constitution as if it is the charter of Australians’ “equality.” The clash of Yes and No became a contest over the kind of “equality” the constitution must secure. But Australians, writes Brennan, are divided in their understanding of the central political and legal concept of “equality under the law.” The referendum result, he writes, “demonstrates that most Australians are yet to accept that there is a case for constitutionalising any collective Aboriginal entitlement, not even the modest entitlement to have a Voice making representations on matters relating to them [emphasis added].”

While some on the conservative side of politics have come to accept collective Indigenous entitlement — Brennan gives the example of former federal MP Fred Chaney’s change of views since the 1980s — many have not. And some who accept legislated Indigenous entitlement still oppose giving it constitutional weight and permanence.

In its most principled form, the Nationals’ objection to the Voice was a refusal to give constitutional expression to a collective Indigenous entitlement. Affirming “No” four months before the Liberals, the Nationals were able to road-test the idea that the constitution would be rendered racist by adding the Voice. In April, when Dutton agreed that Australia needed only a legislated Indigenous Voice, the Coalition could present itself as united in defence of the national “birth certificate” that secures “equality” as a foundational Australian value.


When we consider all of Brennan’s observations together, it is hard to see a moment when bipartisanship was possible. Brennan believes that tightening up the phrase “executive government” might have reconciled Coalition MPs to the government’s campaign. But was this the Coalition’s main concern? In my view, Brennan has underestimated conservative MPs’ antipathy to giving constitutional weight to collective Indigenous entitlement.

That the Coalition no longer contests native title doesn’t mean that all conservatives accept the principle of distinct Indigenous rights. Many continue to regret that native title was foisted on Australia by an “activist” High Court and the Keating government. The Coalition was persuaded by a dogged John Howard that Australia could live with a legislated Indigenous entitlement, and in 1998 he amended the Native Title Act accordingly. But pragmatism’s triumph should not be mistaken for a deeper conversion to other ways of conceding Indigenous rights, especially if that concession is to be in the constitution.

Recovering from the 2022 election, the Liberals under Dutton developed their position on the Voice in the context of a larger question of conservative political identity. A centrist Labor government offering “collective Aboriginal entitlement” — even in the mild form of an advisory Voice — exposed an ideological flank against which it was easy to rally those lacking what Andrew Gunstone calls a “sophisticated” understanding of “equality under the law.”

The Nationals (not least their star Aboriginal senator Jacinta Nampijinpa Price) showed the Coalition how to mine the popular view of equality as sameness and to cherish the constitution as its guarantee. What could moderate Liberals such as Andrew Bragg and Julian Leeser say to counter Price when she presented the Voice as a racist assault on her nation and family?

Brennan’s account of the Liberals is permeated by his ideological affinity with the dissenting moderate Liberals and with constitutional conservatives such as Greg Craven, whose misgivings about voting Yes came to adorn the official No pamphlet. Understating the moderate conservatives’ isolation, Brennan has overstated the consequences of the Labor government’s mishandling of the referendum’s politics.


In The End of Settlement: Why the 2023 Referendum Failed Damien Freeman also tells a story of missed opportunities for dialogue. But Freeman traces the failure back to the Referendum Council, which assisted Indigenous Australians to establish a consensus position (expressed in the 2017 Uluru Statement) but did nothing “to engage with the range of different perspectives that non-Indigenous people brought to this issue.” He refers to constructive interventions in the debate — by policy analyst Kerry Pinkstone and lawyer Louise Clegg — that were not widely discussed. He praises former Liberal MP Julian Leeser and Indigenous leaders Sean Gordon and Mick Gooda for urging the government to reconsider its draft.

Like Brennan, Freeman is highly critical of the Albanese government’s conception of bipartisanship. Eschewing a constitutional convention, he says, the government said, in effect, “This can be bipartisan, if you want to join us in what we’re doing.”

He sees this failure to institutionalise dialogue as symptomatic of the decline of what he calls “settlement politics” — an expectation that political engagement enables people with different programs and values to agree on solutions. In settlement politics, people may have very different reasons for being satisfied with the arrangements they agree on.

To illustrate, he explains how Uphold and Recognise, an organisation he founded with Julian Leeser in 2016, imagined four distinct orientations to constitutional recognition: that it was a “deep moral imperative”; that it would help improve Indigenous Australians socioeconomically; that it would take a vexatious issue “off the agenda once and for all”; that it was the “least worst option” for reconciling Indigenous and non-Indigenous Australians. He believes that political dialogue would eventually have formulated terms of constitutional recognition satisfying — in different ways — each of these four orientations.

Freeman argues that “settlement politics” has been undermined by two developments in our political culture: identity politics and populism. By privileging the experience and perspective of “groups in society that have suffered oppression,” identity politics devalues the perspectives of others and discourages dialogue between those deemed oppressed and their oppressors. “There is no imperative to consider the national interest or common good.”

Populism presupposes that politics is a contest between “elites” (who can’t be trusted because they are out of touch) and “ordinary people” (whose reasonable interests are negated by self-interested elites). If elites have pre-empted the political process, there is no point in “the people” participating in it. Identity politics and populism combine, Freeman argues, to decrease trust in politics and to increase polarisation.

According to Freeman, the Yes campaign became a manifestation of “identity politics.” Many on the Yes side were not disposed to a “settlement politics” approach to constitutional recognition because they were convinced that Indigenous Australians, having experienced colonisation, must be the final arbiters of the terms of reconciliation. Through the Uluru Statement, Indigenous Australians had asked for a constitutionalised Voice, and through the government’s Indigenous working party they had specified certain words to encode its function.

The government’s refusal to enter into dialogue with those who wanted to change those words can be seen as Labor’s capture by “identity politics.” But it can also be seen as Labor’s honouring of the moral authority of the Uluru Statement and its respect for the Indigenous working group, which was the custodian — in effect — of the Statement. What Freeman calls “identity politics” others might admire as empathy for the colonised and respect for the deliberative process that produced the Uluru Statement.

The No campaign, in Freeman’s view, exemplified populist politics. Proponents of the No case presented the Voice as “something designed in Canberra for Canberra people rather than to empower ordinary people on the ground”:

The gist of their argument was that all “ordinary” Australians should send a message that they will not be conned by “the elites.” The people should instead vote NO to a proposition that, they claimed, would divide Australia by race. [The referendum’s] failure, they maintained, would have repercussions for the Labor government, not for the nation.

Freeman thus sees both Yes and No advocacy as making it harder to practice “settlement politics.” He fails to mention Dutton’s August 2023 attack on the Australian Electoral Commission — a dangerous foray into Trumpian populism.


As historian Jeremy Walker argues in his recent article, “Silencing the Voice,” the No campaign can be seen not only as a party-political contest but also, in an international setting, as a successful experiment in conservative populist mobilisation by extra-parliamentary entities.

The No campaign Advance (founded in 2019) is linked by personnel and/or ideology with the Centre for Independent Studies, the Institute of Public Affairs, LibertyWorks, the Bennelong Society (now defunct) and the international Conservative Political Action Conference, writes Walker, all of which are integrated with the global Atlas Network, “an umbrella organisation coordinating 515 ‘public policy research institutes’ across ninety-nine countries.” Campaigning against distinct Indigenous rights aligns with the network’s wider global effort to maintain — in the name of “freedom” — the fossil-fuel industry’s privileged place in the global economy.

Australians were exposed to the latest techniques in political persuasion used by private campaign organisations spawned — or newly inspired — by the Voice debate. The Yes campaign organisations were the Uluru Dialogue, From the Heart (later merging with Yes23), Uphold and Recognise, Liberals for Yes, Australians for Indigenous Constitutional Recognition (the fund-raising vehicle of Yes23), Yes23, and Empowered Communities. The No campaign was waged by Recognise a Better Way, Advance (formerly Advance Australia), Not My Voice, Fair Australia, Australians For Unity, and the Blak Sovereign Movement.

For an account of what emerged in this energised public sphere we can be grateful to media and political science academics Andrea Carson, Rebecca Strating, Simon Jackman, Max Grömping, Phoebe Hayman and Tim Gravelle for their report Influencers and Messages. In the months leading up to the referendum Carson and her co-authors tracked media coverage, Voice-related posts on Facebook, Instagram, X, TikTok and YouTube, and online advertising on Facebook and Instagram. Advance was particularly active on TikTok and Facebook, but Yes23 paid for more online ads (10,218) than all other online advertisers combined.

This prolific messaging cost the Yes and No campaigns quite a lot of money: Yes spent $55 million and No $12 million. Drawing on AEC data, the Sydney Morning Herald reported in April this year that Australians for Indigenous Constitutional Recognition received $47.5 million (but spent $3.5 million less than it collected in donations), of which $2.5 million came from ANZ bank. Australians for Indigenous Constitutional Recognition/Yes23’s total receipts were $10.3 million, including $4.5 million from MYOB and large donations from Westpac, Commonwealth Bank, Wesfarmers, BHP, Rio Tinto, Woolworths Group, Woodside Energy and NAB.

Large donors to AICR/Yes23 included the Paul Ramsay Foundation ($7 million), Sophie Oh ($1.5 million), the Pratt family ($1 million), PwC ($250,000), Mecca cosmetics ($150,000), and the Climate 200 fundraising vehicle ($300,000). GetUp gathered $1.7 million in donations.

On the No side, Australians for Unity received about $11 million, the Liberal Party $1.9 million, and Advance $1.3 million. Other large No donors included Bryant Macfie ($800,000) and Roger Gillespie ($90,000). Clive Palmer spent $1.9 million on anti-Voice ads, and Marius Kloppers gave the No side $100,000.

There was an alternative to this array of donors: public funding. But when parliament amended the referendum machinery laws in March 2023, the government (supported by the Greens and independents) declined to adopt the Coalition’s proposal that official Yes and No campaign organisations be established. Consistent with this position, campaigns were not to be publicly funded, thus avoiding the question — awkward for the government — of whether they should be funded equally.

The government did agree to provide tax deductible gift recipient status not only to organisations campaigning for Yes but also to organisations backing a No vote. But the No campaign subsequently made much of the fact that major corporations were among the funders of the Yes side, and that the Yes budget was much larger.

What did all this money buy? Yes posted more ads on Meta’s Facebook and Instagram than No, and Yes offered a greater variety of messages than No. (Carson and her co-authors didn’t have access to Google’s archive of referendum ads.) No had to achieve majorities in only three states, so it could target those mostly likely to bring a high No vote, while Yes had to win in four states and win the national majority vote. Carson and her co-authors suggest that the No campaign’s success was due to its concentration on two messages: “If you don’t know, vote No,” and “division” would result from “giving Indigenous Australians privileged access to the parliament over other Australians.”

The No campaign ran 280 of its ads under a “Referendum News” banner. Referendum News warned voters of “radical changes to Australian life” wrought by decisions of the High Court, of landowners losing rights, of new payments to Indigenous Australians, and of a change in the date of Australia Day. Some No ads evidenced these warnings by quoting Yes advocates, and some featured Aboriginal advocates of No other than Jacinta Price.

In the new public sphere, social media and “legacy mainstream media” are interdependent. An item in social media could be reported in the press as a newsworthy event, and social media messages often included references to published news items (print or broadcast). Sky News, a bountiful source for social media, also posted prolifically on YouTube, running 490 videos on Voice-related topics from 1 January 2023 to 14 October 2023 (compared with 110 videos posted by Guardian Australia). While parliamentary speech was generally absent from this new public sphere, Sky News’s YouTube posting of Jacinta Price’s speech to the Senate on 16 June 2023 was viewed 441,000 times. No other MP had such reach.

When legislating the referendum machinery in March 2023, some MPs were concerned by the government’s opposition to sending official pamphlets to every household. “We know electoral events are increasingly influenced from social media algorithms,” argued senator Linda Reynolds, but she hoped the pamphlets “would be official material that people can trust.”

The government conceded that a Yes/No pamphlet was necessary. But would it guide voters reliably? Pointing out that the AEC is not empowered to monitor the truth of the Yes and No cases (nor of any political advertising), senator Lidia Thorpe urged — to no avail — that the pamphlet be checked by the Australian Human Rights Commission.

Monitoring information became a task for what Carson and her co-authors call “third-party fact-checkers.” Many stories on social media were subject to scrutiny by RMIT Fact Lab, RMIT–ABC Fact Check and Australian Associated Press. (Disclosure: I was one of the “experts” consulted by these organisations.) The AEC also responded to misinformation and disinformation about the mechanics and procedural integrity of the electoral process.

The third-party fact-checking proved marginal and vulnerable. Among those who did pay attention to “fact-checkers” were those who warned that they were too partisan to be trusted. Sky News accused them of “bias, activism, censorship, targeting the No campaign, and accepting or distributing foreign funding.” The IPA accused them of focusing critical attention on the messages of the No campaign.

The evidence of disputation and conflicting information gathered in Influencers and Messages brings us back to a central argument made in the books and articles reviewed here: that Australians’ thinking about “equality” is divided and confused. According to a certain idea of “equality,” it would have been “divisive” to entrench the Voice in the constitution (and even, in some eyes, in legislation). As Advance’s executive director Matthew Sheahan declared, his organisation’s polling and focus groups identified “division” as a key message: “the big, big factor for people voting No.”

The No forces ran hard with it, and — Influencers and Messages reports — they appear to have persuaded not only conservative voters but also many Labor supporters. “Just twenty-one of Labor’s seventy-eight seats in the House of Representatives returned Yes majorities — and most of those were in inner-city electorates.” In three Labor-held seats the No vote exceeded 70 per cent.

Biddle and McAllister’s survey found that only 62 per cent of Labor voters supported the Voice. In the words of Carson and her co-authors, “Had Labor voters been more united behind the Voice — all other things being equal — the referendum would have succeeded comfortably.” •

Reflections on the Voice: During and After the Campaign
By Andrew Gunstone | Australian Scholarly Publishing | $34.95 | 128 pages

“Safety or Change? The 2023 Australian Voice Referendum”
By Ian McAllister and Nicholas Biddle | Australian Journal of Political Science | Free online

Lessons from Our Failure to Build a Constitutional Bridge in the 2023 Referendum
By Frank Brennan | Connor Court | $24.95 | 146 pages

The End of Settlement: Why the 2023 Referendum Failed
By Damien Freeman | Connor Court | $24.95 | 134 pages

“Silencing the Voice: the Fossil-fuelled Atlas Network’s Campaign against Constitutional Recognition of Indigenous Australia”
By Jeremy Walker | Civil Societies: An Interdisciplinary Journal | Free online

Influencers and Messages: Analysing the 2023 Voice to Parliament Referendum Campaign
By Andrea Carson, Rebecca Strating, Simon Jackman, Max Grömping, Phoebe Hayman and Tim Gravelle | La Trobe University | Free online