Inside Story

Who do you trust?

Early voting, preference dealing, overinterpreted polls and an underwhelming debate: notes from another week of campaigning

Peter Brent 24 April 2025 1492 words

Who was watching, and for what? Anthony Albanese (right) and Peter Dutton during the third leaders’ debate on Tuesday. Alex Ellinghausen/Pool/AAP Image


Early voting has commenced, and by the end of the second day around 1.1 million Australian had already pre-polled. Others would have already sent in a postal vote. With the electoral roll at a touch over eighteen million and official turnout likely to be around 90 per cent, give or take, we’ll probably end up with sixteen million votes, or maybe a few more. At the 2022 election, for the first time, most people had voted by the time the day arrived; that looks likely to happen again.

Early voting is a wonderful convenience, but the downside in this big land of ours is that the smaller the number of election-day voters, the greater the pressure on the Australian Electoral Commission to cut the number of polling places. This, particularly in rural areas, will encourage more postal voting, with all the downsides that involves, namely the potential loss of secrecy (voters could be filling out their ballots in the visual presence of family, friends or anyone else) and late arrivals dragging out the count. Covid-19 caused postal voting to jump from 8.3 per cent of turnout in 2019 to 14.3 per cent in 2022. We await the 2025 figure with interest.

Out in the field, meanwhile — in a move that could have been designed to sink the hearts of Liberal MPs and candidates in teal-ish electorates — the Coalition has struck a how-to-vote-card deal with One Nation. The logic of this move doesn’t immediately jump out: the minor party was already preferencing Libs and Nats ahead of Labor everywhere.

And dramatic reports of One Nation operators tearing up cards that and reprinting them showing Coalition candidates with a “2” in front of them in “key electorates” seems much ado about very little. The old and new cards both had the Coalition ahead of Labor, which barring some other third-party or independent challenge produces the same outcome. It perhaps makes an almost subliminal difference for a One Nation supporter who only glances at it and gets positive vibes about that candidate at number two.

Evidence-wise, the strength of One Nation’s cards too is open to debate: the 2016 federal election, when the party largely preferenced against sitting MPs, provided data in favour of their effectiveness in close contests, but Queensland in 2017 seemed to show much less influence.

It largely depends on how many foot soldiers are available to push the cards into voters’ hands. In 2016 One Nation only ran in fifteen seats; this time (like last) it’s 147. Might we see Coalition volunteers helping out?


There are election results, and then there are the voting-intention opinion polls. Often, big jumps in surveyed support for a major party end up being illusory, dissolving as election day approaches and citizens’ minds concentrate on the more concrete issues that actually decide votes.

The Tampa, children overboard and 11 September 2001 sit in that category: at its peak in October of that year the Howard government had double-digit two-party-preferred leads in the polls but ended up with a modest 51–49 victory. Kevin07 as well: he was massively ahead — sometimes by more than 20 per cent — through the year but support declined during the campaign to a modest 52.7–47.3 win. And the glorious honeymoon for Julia “good day for redheads, Kerry” Gillard in 2010… well, you know how that ended.

The 2020 and 2021 Covid bounces for incumbents, on the other hand, turned out to be the real thing, as Australian state oppositions found to their dismay. In particular, Western Australians and our cousins across the Tasman produced recording-setting re-elections.

Which will the government’s Trump bounce be — fully sustainable or on the soft side? A portion of Labor’s poll lead is generated by the unhinged behaviour of the president of the mightiest, most consequential country on the planet. But our government’s boost is nothing like the one recorded by another centre-left party, the Canadian Liberals, in a country economically intertwined with its huge southern neighbour and literally at risk of invasion. Just a few months ago Canada appeared headed for its routine, roughly once-a-decade change of government election; now a new prime minister and the Trump factor have produced a phenomenal polling turnaround.

Election day in Canada is before ours, on 28 April (their election management body, like ours, has taken lots of early votes). The polls have already tightened somewhat.

I wrote last month that the pollsters’ less-than-stellar recent performances at recent Australian federal elections mean I’m knocking about a point off Labor’s measured primary vote and giving it to the Coalition. If Canada’s Liberals substantially underperform against (poll-driven) expectations on Monday this will be further reason to be a bit sceptical of our own pollsters’ numbers.


Every federal election produces articles and news segments dedicated to the question of whether young voters decide the result? The fact that no one has ever seriously claimed after an election that this actually happened gives a hint of how nonsensical the formulation — applied to any demographic group — really is. Elections come down to everyone, certainly in the contestable seats.

In 2025 the yarn’s silliness has been ramped up by lumping together Millennials and Generation Z (putting an eighteen-year-old first-time voter in the same category as their parents in their early forties) and noting how they now outnumber Boomers! When a panellist during Tuesday’s leaders’ debate said (words from memory) “young voters now outnumber older ones,” it illustrated just how preposterous the narrative is.

The hard fact is that the median age of registered voters increases by about a year every decade. At the close of rolls this year it seems to be about 48.5, give or take.

And while today’s younger voters are clearly less attracted to the major parties, there is no serious evidence to support the notion that they are resisting the pull of political conservatism as they age. This theme mostly comes from the Australian Election Study, whose misuse of data I went into a couple of years ago. And, as we know, young people are less likely to turn out to vote.


Speaking of the debate, I thought Anthony Albanese had a pretty awful time of it on Tuesday evening. The camera set up, which shrank him to appear about two-thirds the size of his opponent (for the record it’s 5’11” versus 6’1”) did him no favours, but he also spent quite a lot of time squirming and shifting in contrast to Peter Dutton’s solid stillness.

Watching the PM reminded me of John Howard’s demeanour in all his debating efforts as incumbent, 1998 to 2007: timid and, well, a little shrivelled. As the record shows, there’s no reason to believe these debates make more than a negligible difference to election results. Nor have I prioritised the actual content of what was uttered, because people don’t really do that either, and in my opinion the impressions formed by visuals and sounds are more important to the unattached voter. If any were watching.

For his part Dutton had evidently taken notice of urgings along the lines of “You’re funny in private Pete, show your personality.” But the “couldn’t lie straight in bed” jibe, obviously planned, fell rather flat for all but the true believers. Commentators didn’t find it as electrifying as when Mark Latham wielded it against Howard in 2004. That was game on! That same prime minister went on to develop a catchphrase that was really just saying out loud incumbents’ quiet message at all elections: okay, there are things you don’t like about me, but at least I’m a known quantity, you’ve seen me operate the levers of power, and who do you distrust least to do so after election day?

A lesson Dutton could have taken on board from 2004 is that it’s fine for an incumbent to be seen as a “conviction” politician, but a “conviction” opposition leader can be discomforting to the voter.

There was a time when all prime ministers routinely accrued that tough persona. They came in, broke promises and implemented policies that turned out to be not so painful after all, at least for the vast majority, and voters gave them points for doing what was right for the country. Thank you for making me eat my greens.

This is not a dynamic the current prime minister enjoys, but then neither did his recent predecessors. Rudd was probably the first post-conviction prime minster, and while Tony Abbott gave it a red hot go, he illustrated that this politics business isn’t as easy as it looks. But we also shouldn’t forget that the post 2005–08 Senate numbers have been difficult for all governments. This fact, unappreciated by commentators who loudly gnash their teeth about the lack of reform, renders it all but impossible to get anything bold through parliament — so why would a politically astute prime minister even try?

Whoever forms government next month will face another difficult upper house. People might want to stop complaining and look at reforming our electoral system. •