Josh Frydenberg Scott Morrison Budget 2019 federal election
(Image: AAP/Lukas Coch)

As the election nears, the attention paid to polls will inevitably ramp up, starting with another round of polls next week ahead of the budget and what will become the start of campaigning for the May election.

In the rush, some basic rules about interpretation of polls will be ignored even more comprehensively than usual. Small shifts in primary votes within the margin of error will be taken as actual, real shifts in voting intention. And journalists charged with writing polls up at both News Corp and Nine will fall victim to the post hoc ergo propter hoc fallacy of attributing those shifts to events that dominated the political narrative in the days leading up to the polls.

It’s not just journalists who fall victim to this fallacy; politicians do too. Malcolm Turnbull explicitly claimed that Tony Abbott had sought to undermine him by timing his disruption to coincide with the lead-up to Newspoll; he even took to announcing his own perceived vote-shifters, like the Snowy Hydro thought-bubble in the days before Newspoll to counter it. But it’s a joint delusion of Press Gallery journalists and politicians that the majority of Australians are paying the slightest attention to what’s happening in Canberra on a day to day basis.

Most couldn’t give a damn; many actively loathe politics (and journalists). The idea that everyday Canberra brawling shifts voting sentiment on planet Earth has little evidence.

Sign up for a FREE 21-day trial and get Crikey straight to your inbox

By submitting this form you are agreeing to Crikey's Terms and Conditions.

A classic example of this is the fate of Nine’s dreadful Ipsos poll, which is run too rarely to be useful and has a puzzling flaw in its methodology that has persistently inflated the Greens vote. You’ll recall that after the passage of the medivac legislation, the Press Gallery was abuzz with talk of Scott Morrison’s “Tampa moment” and how it would provide the basis for a comeback election victory. And Ipsos appeared to endorse that with a poll a week later that had Nine journalists declaring “Labor has suffered a sharp fall in popular support after a week of incendiary political claims over border protection.”

Except, it hadn’t suffered anything of the sort. The result was being compared to a previous poll before Christmas, and it was demolished a few days later by Newspoll, which showed, to the dismay of Simon Benson and co at the Oz, no shift at all since the previous fortnight. 

The soundest conclusion to draw was that most voters were entirely indifferent to the medivac issue that had gripped politicians and journalists and become a totem for both left and right. And that’s terrible news for everyone involved in political journalism, because the last thing any of us want — no matter where we work — is the people who employ us thinking that most Australians have little interest in what we breathlessly report.

The experience of recent years suggest that polls move in two ways. They turn slowly in one direction or another in response to a disengaged electorate’s vague impression of what’s happening in politics. Malcolm Turnbull slowly turned around his government’s polling from a big Labor lead to, more or less, 50-50 (he was very unlucky never to benefit from a one-off 51-49 or 52-48 in his favour due simply to statistical noise). Strong jobs growth, the ouster of Barnaby Joyce and a few months relatively crisis-free appeared to do the job for Turnbull, until he mismanaged expectations around the byelections last year and the right seized its chance to kill him off.

The other way — as that event also demonstrated — is if there’s a truly seismic event in Canberra that pretty much everyone tunes into, regardless of what they think about politics. The ousting of a Prime Minister — despite its frequency — is still such a seismic event. And that sent the Coalition’s stocks plummeting when they ditched Turnbull, in contrast to the huge surge to the Coalition when Turnbull replaced Abbott.

With around six, or maybe seven, weeks to go ’til polling day, Scott Morrison doesn’t have the time for a slow turnaround of the kind Turnbull accomplished. He needs a seismic jolt, and the budget is his last chance to do it — the budget is one of the few set-piece political events that voters tune into. Problematically, budgets have also proven to be relatively poor at delivering any benefits to incumbents — few political journalists now bother referring to a “budget bounce” since the evidence against such a thing is so extensive.

That means Josh Frydenberg will have to produce something stunning on Tuesday night, something that can change the minds of the large numbers of voters who will be paying at least momentary attention. It can be done, especially given he will have lots of revenue to splash around. But he’ll also need to hope Pauline Hanson/Barnaby Joyce/a disgruntled cabinet minister doesn’t wreck the whole thing by distracting everyone.