Everything pointed to a Labor win in the 2019 election. But it is only a pretence that we are in the certainty business
The
federal election wash-up is a quaint sort of ritual. There are,
inevitably, explosions of feelpinion about why one side won and the
other side lost. The victors rush to write the first draft of history,
and on the losing side there is a frenzy of backside covering and
jostling for future advancement, masquerading as deep insights about who
failed to do what and when.
All of it, the initial download – masters of the universe versus the vanquished – is inherently unreliable because the protagonists are unreliable. They are still processing what has happened and they are positioning for the next thing, because that’s how the psychology of politics works. Everything is refracted through post-campaign decompression and self-interest.
In the category of much more worthy than playing stenographer to a group of people suspended after a visceral experience, in states of ebullience, depression or internecine scheming, are the efforts to find some kind of factual underpinning for post-election analysis. Sadly though these undertakings, while earnest, also have their limits.
We can look at what the data tells us, and Guardian Australia’s terrifyingly clever Nick Evershed did just that. Nick’s big data crunch
showed electorates that swung harder to the Liberal and National
parties were more likely to have higher unemployment, lower income,
lower levels of education and fewer migrants, and the electorates that
swung to Labor were more likely to have higher levels of education, more
young people, more people in work or study, and more people over the
age of 80.All of it, the initial download – masters of the universe versus the vanquished – is inherently unreliable because the protagonists are unreliable. They are still processing what has happened and they are positioning for the next thing, because that’s how the psychology of politics works. Everything is refracted through post-campaign decompression and self-interest.
In the category of much more worthy than playing stenographer to a group of people suspended after a visceral experience, in states of ebullience, depression or internecine scheming, are the efforts to find some kind of factual underpinning for post-election analysis. Sadly though these undertakings, while earnest, also have their limits.
This is all pretty interesting, but fortunately for all of us, Nick is the sort of practitioner both brainy enough and respectful enough of his audience to own and telegraph the limits of this reportage. He told readers very clearly we don’t yet know why more people voted for the Coalition over Labor – that being the truth. A subsequent piece from Shaun Ratcliff, a lecturer in political science at the United States Studies Centre at the University of Sydney, picked up Nick’s theme, pointing out that we cannot make inferences about individuals from aggregate, electorate-level data. We’ve got to look a bit deeper.
So where I am going with all this? It’s a preamble to expressing one of the most dangerous thoughts in political journalism – a thought we don’t express often enough between ourselves as professionals, and almost never in public.
Here’s the thought: we don’t know, with perfect certainty, what is happening and why.
I’ll say it again. We don’t know.
This is not me saying we political reporters, as a cadre, lack expertise. The tenacious souls amongst us hang in there for decades, studying our ecosystem with great diligence and care. We learn about the politicians we report on, observe their habits and rituals, building up a level of insight that informs our reporting and commentary – this vocation being a life’s work for most of us rather than a diversion on the road to somewhere else.
But we don’t have a crystal ball and we don’t have perfect wisdom.
It’s funny. I had high hopes when we reached the end of the “voice of God” era in journalism, the era when the great men of newspapers delivered righteous pronouncements and the mortals nodded, absorbing their stupendous wisdom, at the time when the disruption of the internet and the rise of the reader exposed that whole Ponzi scheme for the self-absorbed pomposity that it was – I hoped we’d be liberated from the tedious tyranny of being self-appointed sages.
But the occupational requirement to be predictive has simply morphed into a new phase. Now the 24/7 environment has spawned the flourishing chat shows on the rolling news channels that require everyone to keep talking, preferably in definitive terms rather than boring people with grey area.
So much talking, filling up the dead air that swells dangerously when the drama of politics is insufficient to sustain the next 10 minutes. Because of that relentless requirement to be predictive, journalism has resumed the collective pretence post-disruption that we are in the certainty business, when the business we are actually in is the uncertainty business.
This most recent election campaign is an obvious case in point. Three years’ worth of opinion polls, seemingly the most solid of trends, pointed to a Labor victory on 18 May. That evidence base, coupled with Labor victories during the byelection season triggered by the section 44 fiasco, shaped everyone’s perceptions of the contest – both the protagonists and the chroniclers. That, and suboptimal private polling, created a perception frame so rigid that senior people inside the Labor campaign did not know they were losing, a development both extraordinary and true.
There are a couple of potential explanations for what happened. Perhaps the evidence base was right within the margin of error over the past three years but wrong in the campaign, failing to detect a late swing to the Coalition. That’s possible.
Or perhaps this was the campaign where two significant things happened: where fake news became a serious feature of an electoral contest for the first time (this subject is worth a whole other column or piece of reportage), and underlying problems with polling surfaced in a way that’s now impossible to ignore – problems such as the difficulty of getting representative samples now people routinely screen calls to their landlines, which Peter Lewis from Essential has alluded to.
It’s not just the fact that a number of polls were clearly problematic, either for a long time or a short time. The way polls are reported also plays into the over-egged predictive imperative I’m decrying this weekend, with margin-of-error movements dressed up as profound developments directly connected to the intrigues in Canberra, rather than blips or noise.
In this spirit, it has been funny to watch some of the post-election journalistic castigation of the Labor campaign on the basis that it failed to grasp the zeitgeist, as if this was some localised or situational form of idiocy, when the truth is the great bulk of us failed to see it coming.
On the afternoon of the election, in preparation for the night’s coverage, I spoke to a couple of government players who reported to me that Scott Morrison was really upbeat, and “Hirsty” (the Liberal campaign director Andrew Hirst) was similarly pumped – but the people I spoke to didn’t really believe it. They had no idea how the night would end – and both feared not well for the Coalition.
The good news is there are lessons to be learned from 2019 if we are humble enough to learn them. The principal lesson is whatever the ubiquity of the apparently objective pseudo science, don’t fall into a perception trap so profound that it becomes an implicit prediction. The secondary lesson is embrace the fact that politics and journalism are the most human of businesses, and where there are humans, unpredictability and uncertainty is always part of the deal.
No comments:
Post a Comment