If the biggest surprise of May’s federal election was its lopsided result, a secondary one was the extent to which the Coalition had guzzled down the Voice referendum Kool-Aid. Down to their bootstraps, it seems, those campaigners really, truly believed that the 60 per cent No vote two years ago held great promise for the general election.
In truth, as at all our earlier forty-four constitutional referendums, the question of who would govern the country for the next three years was not on the ballot. And as with most of the others, the Voice question was a tenth-order issue for probably 90 per cent–plus of electors. It joined the thirty-six previous defeats, and its size was also unexceptional in the context of Labor governments’ midterm attempts at constitutional change. (As it happens, governments have nearly always won the election following a referendum loss.)
It’s not clear whether Liberal hardheads saw 14 October 2023 as an indicator, like a large opinion poll, or as a process that would shape voter behaviour, so that the campaign and act of rejecting the government’s proposal made Australians more pro-Coalition.
For whatever reason, throughout the 2025 election campaign Peter Dutton repeatedly brought up either the Voice or the broader topic of Indigenous Australians, with a particular focus on violence in Alice Springs and “Welcome to Country” ceremonies. Of the latter, the opposition leader opined repeatedly that there were too many, and under his government there would be far fewer. In one televised debate Dutton demanded that Anthony Albanese provide similar clarity; the prime minister expressed support for the ritual with his characteristic bland ineloquence. (Most “Welcome to Country” ceremonies actually have nothing to do with the federal government.)
As with the Voice (in the end), it’s likely that most voters agreed with Dutton — but so what? Election campaigns routinely showcase fluff and trivia, but deep down voters are mostly motivated by livelihoods, security, putting food on the table, particularly in a cost-of-living crisis. Still, in the campaign’s dying days the News Corp–Coalition tag team, citing words from Penny Wong, ran hard on a re-elected Albanese holding another Voice referendum. It was perhaps the feeblest “scare campaign” the nation has ever seen.
(The foreign minister’s formulation — “We’ll look back on [the Voice] in ten years’ time and it’ll be a bit like marriage equality” — made little sense anyway. Same-sex marriage was a global trend whose time most Australians finally believed had come, and the voluntary, indicative quasi-plebiscite had nothing to do with the Constitution. A Voice, by contrast, could simply have been legislated but its designers wanted it constitutionally enshrined, which all but doomed it. The chances of the Constitution ever containing an Indigenous Voice to Parliament are now approximately zero.)
The Liberals’ pollsters, Freshwater, went even further, using their respondents’ recalled Voice vote to second-guess their stated 2025 voting intentions. Or so it seems.
The day after the election, Freshwater’s boss laid out in the Australian Financial Review (for which the pollster had also provided survey outcomes for several years) three reasons why their final voting-intention numbers, whether published in the AFR (Labor ahead 51.5–48.5 per cent, way off the actual 55.2–44.8 result) or provided privately to the Libs, got it so wrong.
“Polling appears to have overestimated Labor ‘defectors’ to the Coalition,” ran his first justification. “Particularly, those who voted No at the Voice referendum. Early indications suggest that ‘Labor-No’ voters just didn’t switch over to the Coalition in the big numbers estimated.” (Italics added.)
“Estimated” is doing lots of heavy lifting here. How did they do it? Did they presuppose a proportion of those who said they’d voted No in 2023 and would be voting Labor in 2025 (or had already voted early) would actually vote for the Coalition instead? Or some variation of this? It makes no sense, and the appropriate word would have been something like “assumed” or “hoped for.”
The October 2023 vote was of course not about some arcane, dry piece of law, but about Indigenous Australians. Had it been instead about recognising local government (a chestnut attempted in referendums in 1974 and 1988 and planned for one in 2013 until bipartisanship collapsed) and had it been similarly walloped (pretty much a certainty — all midterm Labor referendums have suffered that fate) would the Coalition have invested it with so much meaning?
Probably not. Topics involving minorities hold a special place in our country’s electoral narratives. Ever since the 2001 “Tampa election,” the conviction that what are politely called “values” can rescue the Coalition at the ballot box has taken a firm grip despite all the contrary evidence. John Howard himself attempted unsuccessfully to resurrect that 2001 dynamic in his final 2007 hurrah as prime minister, first with the Northern Territory intervention and then with African immigrants.
Dutton rose to the peak of the Liberal Party by embodying this particular element of the Howard legacy. The socially conservative Liberal leader who connects with blue-collar voters by sympathising with their resentment at elites, multiculturalism and supposed special rights for racial minorities remains an alluring tale. There was no great demand from the electorate for Dutts to be leader, but he had long been a big hit in the conservative “base” and among the blathering heads at Sky After Dark. Throughout the 2022–25 term he repeatedly, often with grating gratuitousness, reached for the immigration lever, perhaps as red meat for the base, perhaps to connect with less-engaged voters, perhaps to make himself feel good, or perhaps all three.
That was presumably what drove him to fleetingly ruminate in March this year on a constitutional referendum (!) to enable the government to deport criminals who had dual citizenship.
Once again, most Australians would doubtless instinctively support the idea, but it is tangential at best to what they look for in a prospective government. That legendary 1992 Bill Clinton war room Post-it note, “It’s the economy, stupid,” could do with elaboration: “It’s economic security, and security generally, for people and families.” As we’ve seen recently in other advanced democracies, fears about out-of-control immigration, especially by people who look and act differently, can slot into the “security” framework, but Australia simply lacks anything like the required numbers. Similarly, not-so-subtle hints about special rights for Aboriginal Australians can get traction in a standalone vote on a particular topic but are barely of interest at a general election.
Dutton’s repeated forays into “values” were out of place for voters looking for solutions for Australia’s years-long per capita recession. Much worse, they provided another reason to be wary of his Trump-like tendencies.
And what possessed his party’s pollster to apparently overlay their results with “estimates” of large numbers of “Labor-No voters” switching to the Coalition? •
