Will the ghosts of 2019 return to haunt the 2025 election?

The opinion polls predicted a Labor victory in 2019 but got it horribly wrong, and the election was instead won by the Coalition. It’s unlikely to happen again.

Election day has finally arrived and as the final votes are being placed into ballot boxes all around the nation, the last round of opinion polling has confirmed the consistent message that has been appearing during this campaign: Labor is set to win the 2025 federal election.

Whether it secures a majority or falls just short and governs in minority will be confirmed late on Saturday night, but across the major pollsters – Newspoll, Resolve, Essential, Morgan, YouGov, and Demos – the two-party-preferred range sits between 52–48 per cent and 53–47 per cent in Labor’s favour. There’s been no late swing, no surge of momentum for the Liberal–National Coalition, and certainly no tangible sign that a ‘miracle victory’ is brewing but despite this consistency, the ghosts of the 2019 federal election still hovers in the background of this election.

In 2019, opinion polls predicted a Labor victory under Bill Shorten. Instead, Scott Morrison stunned observers with a narrow but decisive win, proving just how wrong the polling industry can sometimes be. But there are a few critical differences between 2019 and 2025 – differences that reduce the changes of another upset result.

The most critical issue is the campaign performance itself – leader of the opposition Peter Dutton has run what has been widely regarded as a disastrous campaign: reactive, negative, undisciplined, and disconnected from the political centre. In contrast, the Prime Minister, Anthony Albanese, has run a steady, if unspectacular, campaign defined by competence and clarity. There have been no major scandals, no self-inflicted wounds – except perhaps for a momentary media frenzy over Albanese’s minor fall at a public event. If that was the worst controversy of Labor’s five-week effort, then it’s a sign of a campaign that stuck to the plan and delivered the message.

Polling firms, too, have learned from their mistakes. After the methodological failures of 2019, many recalibrated how they sample, weight, and analyse voter sentiment – particularly around undecided voters and disengaged demographics. While pollsters remain cautious and always warn that their polls are not predictions but snapshots at a given time, they also acknowledge that consistent polling across all firms this close to election day usually signals a reliable outcome – at least in terms of the likely winner, if not the precise seat tally.

Crucially, Peter Dutton is not Scott Morrison. He enters the final week with a net approval rating of minus 24 points, far worse than Shorten’s minus 8 points at the same stage in 2019. Albanese’s approval rating is currently +1 point – identical to Morrison’s rating in the lead-up to his surprise victory. For Dutton to replicate that 2019 ‘miracle’ victory, he would at least need to be in Albanese’s position. Instead, he is in a far worse place than even the losing candidate from 2019. Personal unpopularity matters, particularly in tight contests where undecided voters often look to leadership qualities as a deciding factor.

Another structural disadvantage for Dutton is that he is trying to win from opposition. The political fortunes tend to favour incumbent governments, especially first-term ones, unless there is widespread dissatisfaction. But Albanese’s administration, while not immune to criticism, has largely avoided scandal, has delivered economic stability, and has presented a unified front. Cabinet unity has been tight; while there was a cabinet reshuffle in 2024, no ministers have resigned in disgrace; no leaks have derailed the campaign. The same cannot be said for the Liberal Party, where internal players continue to leak stories to the media and, in some cases, directly to the Labor Party itself.

While both major parties have brought out the ghosts of 2019 in different ways – Labor as a cautionary message against complacency; the Coalition as a call to faith and belief in electoral miracles – the context has changed and the political landscape is now quite different. The leaders are different. And the electorate is different. Voters are now more engaged in the pre-polling process, with nearly half the electorate already having cast a ballot before election day. That shifting dynamic makes last-minute momentum harder to generate, and harder still to measure.

It’s also worth noting that the 2019 opinion polling errors were not as catastrophic as often portrayed – while the figures were incorrect, the result was still within the margin of error, but the opinion polls miscalculated where that error would land. This time around, every major pollster is showing a clear Labor lead – albeit a narrow in some cases. For all of them to be wrong, again, and in the same direction, would represent an extraordinary statistical anomaly.

In the end, elections aren’t won or lost by opinion polls, but they do provide a barometer of where the public mood is heading. And right now, that barometer points toward a second term for Anthony Albanese. If Peter Dutton were to win from here, it would not simply be a comeback – it would be a political resurrection of unprecedented proportions. But the numbers don’t lie: an unpopular leader, a faltering campaign, a fractured party, and a hostile public mood all point in one direction. Unless something extraordinary occurs at the eleventh hour, the Coalition’s hope lies not in government, but in saving enough of the political furniture to rebuild – and, more than likely, under new leadership.

The final predictions: A likely Labor victory, a fractured Coalition, and a strong crossbench

After five weeks of campaigning, political theatre, strategic missteps, and moments of occasional clarity, we reach the most important part of this campaign: our prediction for the result of the 2025 federal election. And while it’s always wise to be cautious – especially with the spectre of 2019 threatening to disrupt these predictions – the political landscape, polling data, and the campaigns themselves all point to one likely result: a Labor victory, most probably in majority government.

There are many reasons for this prediction, and most of them are rooted not in soaring support for Labor, but in the dismal performance of the Coalition. Albanese’s government, while far from perfect, has been stable, competent, and relatively scandal-free. The public perception of Labor has been bolstered by a sense of steadiness, economic management, and ministerial unity. Even its political failings – such as its lacklustre housing policy, an underwhelming anti-corruption commission, and its cowardice on the genocide committed by Israel in Palestine – have not translated into a widespread voter backlash. These issues may weigh heavily on progressives and political observers, but for most of the electorate, they haven’t defined the government.

On the other side, Dutton’s campaign has been relentlessly negative, often chaotic, and completely devoid of compelling vision and the Coalition has failed to articulate a clear reason for voters to turf out a first-term government. And to run a negative campaign successfully, a party must be disciplined, precise, and highly strategic – none of which describe what we’ve witnessed from the Coalition over the past five weeks. Simply turning up during a campaign with a relentless megaphone of negativity rarely wins an election: voters need a positive light and direction for where the country is heading over the next three years.

The late release of the Coalition’s policy costings – on the penultimate day of the campaign – reinforced this sense of disorder and gave the media and electorate no time to evaluate them, with just the vague promise of being “$10 billion better off”, in conjunction with a final photo-op at the petrol bowser to push a promised 50 per cent reduction in excise. Petrol prices have actually fallen by around 15 per cent since the last election, this was a misstep – yet another sign of a campaign looking for quick stunts rather than serious substance.

Dutton’s personal unpopularity has dogged him during the entire campaign and he is the least liked major party leader in polling history heading into a federal election. His only serious political achievement over the past three years – the defeat of the Voice to Parliament referendum in 2023 – may have elevated his standing among conservative die-hards, but has left him politically toxic in much of the country.

The broader strategic failure is evident in the leadership itself and even the Liberal Party probably realises that Dutton should never have led them into this campaign. Yet he survived not because he inspired confidence or electoral momentum, but because there was simply no one else with enough credibility to replace him. Names like Angus Taylor or Andrew Hastie have been floated, but none present a serious path forward for a Coalition that has lost its ideological centre and strategic coherence.

So where does that leave us?

The Coalition cannot realistically reach the 76-seat threshold needed to form a majority. It is equally unlikely they could form a minority government. That path would require a miraculous confluence of preference flows, independent crossbench support, and a primary vote that simply isn’t there. While isolated upsets might occur on election night, the idea of a Dutton-led Coalition negotiating a path to government in this political climate borders on fantasy.

In contrast, the independent and minor party vote continues to reshape Australian politics. The crossbench is likely to remain historically large – even in the event of a Labor majority. Most community independents are expected to retain their seats, particularly those in affluent, formerly Liberal electorates. The teal independents who defeated Morrison-era conservatives are unlikely to see voters return to the even more regressive Peter Dutton. In fact, community independents might be able to gain ground – with likely pickups in seats such as Bradfield and possible upsets in Labor-held seats such as Fremantle.

The Australia Greens could also increase their lower house representation, although whether an expected higher primary vote translates into seat gains remains uncertain. Their vote could actually rise nationally but could also concentrate in places where they’re already strong, yielding little net change. Still, they remain an important part of the crossbench and a likely influence on Labor policy in a second term.

So, here are our final predictions:

• Labor to win government, most likely in majority, but a minority government still remains a possible outcome.

• The Coalition to lose seats overall, with the Liberal Party further weakened and facing an existential leadership crisis.

• Peter Dutton to remain leader briefly post-election, but face immense pressure to resign within days or weeks.

• The crossbench will remain large, potentially growing to 19–21 members, including teal community independents, Australian Greens, and other independents.

• No viable path to victory for the Coalition, despite some public bluster and last-minute optimism from a handful of Liberal Party MPs.

• Significant long-term implications for the shape of Australian politics, with further erosion of the major party duopoly.

Australians have sent signals in recent elections that they want better representation, more integrity, and less of the tribal culture war politics that defined the Morrison and now Dutton era. If the outcome reflects what the polls and the campaigns suggest, this election will confirm that the public is still on that trajectory – and that the Liberal Party, as it is currently constituted, is not part of that future.