‘No body saw it coming,’ the mainstream media repeated endlessly when results emerged on Saturday evening 18 May, contrary to all of the predictions of the polling companies, even last minute and exit polls, the betting markets and a vast army of commentators, Alan Jones being a notable exception. Just like the 1999 republic referendum, Brexit and the election of Donald Trump.
No one saw it was coming?
Not so. Early on Wednesday morning before the election, The Spectator Australia carried my prediction, supported by an analysis from a table of 31 seats, that the Shorten Labor Opposition would be defeated and the Morrison government returned.
This coincided with an interview with the very wise young commentator Michael McLaren on 2GB, 4BC and across the Macquarie Media Network. On that I recently met a very smart intense young man in Hahndorf in South Australia where I was to chair a candidates’ forum. He said he heard me regularly on 2GB. I asked which station broadcasts that into South Australia. ‘None,’ he replied. ‘I hear you by podcast.’
My seat-by-seat analysis was of course not perfect, but in aggregate it proved to be correct. In retrospect, I should have added a column indicating the concentration in each electorate of the gullible and virtue-signalling. Warringah would have led, with Wentworth second.
This incidentally reflects the 1999 referendum where such electorates were among the few which voted for the politicians’ republic. They mindlessly endorsed the removal of significant checks and balances on the politicians, endorsing the principal reason they advanced for change, the blatant lie that we don’t already have an Australian as head of state.
This comment is not meant as a criticism of pollsters, but of using polling as an infallible source of truth, rather than as the useful tool it can only be. It should especially be complemented by a good dose of common sense.
This has been in my mind ever since, as a boy in 1948, I saw how the polls had been similarly worshipped in the US so much so that the Chicago Tribune prematurely and erroneously published a report under the headline DEWEY DEFEATS TRUMAN. This is like the Australian betting company which paid out on Labor winning days before the election. They would have had to pay out the real winners on Sunday. Indeed, one Speccie writer has told us the amusing story of a man who was paid out his winnings for backing Labor on the Friday, and promptly went and re-bet them on a Liberal victory. A veritable win-win.
But back in the US in the Forties, counting soon demonstrated the opposite to the Tribune headline. The truth was that President Truman defeated the favorite, Governor Dewey.
Later, when I was chairing the engine room for the No case in the 1999 republic referndum, the polls usually reported a higher vote for a republic than transpired. This complemented a vigorous, often no-holds-barred campaign by the mainstream media (Alan Jones again a notable exception) that a republic, by which they meant any fake politicians’ republic, was inevitable.
This was repeated again in the US presidential election when the mainstream media were an integral part of the Clinton campaign. Among a stable of pollsters, only Rasmussen and the LA Times suggested a result close to that achieved.
It’s surely time now to stop this mindless adoration of the polls as if they were the Pope speaking Urbi et Orbi. They are a useful tool but only a tool. Not as they have become, part of political campaign weaponry as they were so notably used to bring down Tony Abbott in a pincer movement with selected commentators and journalists fed by leaks from the treacherous within the cabinet.
The Australian people, by their decision on 18 May, have dethroned the polls and the commentariat as the arrogant rulers of our Commonwealth. The nation has declared, ‘The emperor has no clothes!’.
The principal defect in the election polls was the assumption pollsters knew how voters for small parties, 15 per cent nationally, would allocate their preferences.
The method chosen was hardly scientific. Most used preferences from the 2016 election. But common sense would tell you circumstances are significantly different today than in 2016. Almost alone, the Morgan Poll asked, say, One Nation voters how they would allocate their preferences. But this would involve sub-samples (say 40 in 1,000) which were too small to be meaningful.
So I summarised the differences between the 2016 and the 2019 elections as affected significantly by three ‘TPD’ factors, piquantly named Thief, Palmer and Del-Con.
You’ll have to go to the piece to find out what the last two named mean but the Thief factor was so called because of the threatened theft of tax refunds on franked dividends but also covered all increased taxes and costs.
In separate columns I set out two crucial statistics as they applied to each seat, franked dividends and negative gearing. These were to show in each electorate how a significant number of voters, either directly or through preferences, would vote.
Comparing this with the margin with which each seat was held, I then made a common sense assessment as to what would happen in each seat, whether Coalition or Labor.
Take for example Herbert, held by Labor on an 0.02 per cent margin with Newspoll reporting 28 per cent would vote for small parties. There are 9,532 families involved in negative gearing no doubt outraged over Labor’s property taxation changes.
That is probably 20,000 voters. I concluded that, with disaffected ‘tradies’ and farmers, Adani, self-funded retirees, the religious, free speech advocates and so on, this should swamp the margin and hand the seat to the LNP. Phillip Thompson won the seat from Labor; 57.6 per cent to 42 per cent with a swing of 7.6 per cent.
Applying common sense to all this information, I concluded the Coalition should do better than either the polls or the betting markets were saying. They would hold most of their seats and win more.
Which is precisely what happened.
Got something to add? Join the discussion and comment below.
You might disagree with half of it, but you’ll enjoy reading all of it. Try your first 10 weeks for just $10