Crying Wolf: More Embarrassingly Bad Tabloid Poll Reporting

 "Anthony Albanese could be on track to being a one-term Prime Minister, with a new poll showing Labor's primary vote crashing in three major states.  The federal government is in serious trouble in the eastern states - where most of the seats are - with Labor down to 24 per cent in Queensland, 28 in Victoria and 32 in NSW."

"Labor’s primary vote has crashed to just 24 per cent in Queensland, 28 per cent in Victoria and 32 per cent in New South Wales, the wolf + smith shows.  But Labor is dominant in South Australia, where its primary vote is 41 per cent, and 60 per cent on a two-party-preferred basis.  The poll – which measured both state and federal voting intention – suggests the government is in dire trouble in the eastern states, with just 43 per cent of the two-party-preferred vote in Queensland and 48 per cent in Victoria."

This is how the Daily Mail and the Murdoch tabloids (Courier Mail/Daily Telegraph etc) respectively wrote up a massive new poll by an initially mysterious outlet wolf + smith.  But this was in fact another example of laughably incompetent poll reporting from these outlets, one that again happened to be in service of the narrative their right-wing readers would want to see.  What the poll in fact found is very different.  The state-level figures these outlets were commenting were state voting intention not federal.  This was made so abundantly clear in the poll report that, among other subtle hints, the whole of page 10 of the poll report is devoted to making it clear that the rest of the report is state not federal.

Page ten looks like this:

This whole page is apparently, somehow, not large enough or clear enough to make it clear to journalists from the Dailys Mail and Telegraph that what follows (in sections helpfully also introduced as "nsw politics", "victorian politics" etc) is state voting intention.   Abysmal!  There are even more helpful pointers like " IF A NSW ELECTION WAS BEING HELD TODAY AND YOU WERE TO VOTE, WHICH OF THE FOLLOWING PARTIES WOULD RECEIVE YOUR FIRST PREFERENCE VOTE?" to indicate that this is state voting intention.  The Murdoch tabloid article even manages to realise that the poll canvasses both state and federal intention and still somehow represents the state figures as federal.  Unbelievable!

What the poll actually finds concerning federal voting is rather different.  The poll finds primary votes of Labor 29% Coalition 36% Greens 13% One Nation 6% Independent 11% Others 4%.  The two-party preferred is 51-49 to ALP by "past preference flows".  As far as the poll report goes is "with Labor’s majority in danger."  Majority in danger, it says (and fair enough), not government.     

There aren't any specific state primary or 2PP breakdowns provided, but what the report does provide is state by state primary vote gaps between the major parties.  I provide these with the 2022 results in brackets below.  

SA ALP +4 (-1.08) 
Tas ALP +2 (-5.68) 
WA ALP - 3 (+2.06)
Vic ALP -7 (-0.23)
NSW ALP -9 (-3.16)
Qld ALP -10 (-12.22)

This means the poll actually finds substantial swings to Labor on major primary gap in SA and Tas, and a slight swing to Labor in Queensland, with swings of a few points away (note again: this is major party gap, not 2PP) in WA, Victoria and New South Wales.  (The change in gap is halved to get the swing.)

If I treat the primary gap swings as 2PP swings and apply them to the expected post-redistribution boundaries, by uniform swing by state Labor would lose Aston (if counted off the 2022 baseline), Bennelong, Gilmour, Robertson and Paterson (maybe Chisholm as well depending on whose estimates you use) and gain Sturt and Bass.  However, the poll had swings to the Greens and independents and away from the combined One Nation/UAP, so that 75-seat projection (minus any seats dropped in NT) is a pessimistic reading of the poll for Labor.  Most likely on these state numbers Labor would win a majority if its preference flows held up from 2022.   That's a big if at this moment according to all three pollsters that are running respondent preferences, but the state numbers are hardly Dutton-in-the-Lodge territory.

In particular, there's a big difference between the major party vote gaps in Victoria (federal 7 points, state 12 points) and Queensland (federal 10 points, state 18 points) that shows why the poll looks so bad for Labor if one confuses its state results with the federal picture.  

Even if one sees a 10% primary vote gap in "marginals" in this poll, should one be alarmed on Labor's behalf?  Not really, because the average gap in marginal classic seats last time was 6.8% and the gap swing is about the same as in the poll overall.

As for the mysterious wolf+smith with its enormous sample size for a federal poll, it turns out this is an outfit directed by Jim Reed of Resolve and carrying similar hallmarks including a high Independent vote and a relatively low Labor primary (more or less cancelling out on expected 2PP).  Former Scott Morrison advisor Yaron Finkelstein is also involved.  However there is no mention of who is involved on the wolf+smith website (did the Murdoch tabloids even know this was an outfit that involved a Ninefax pollster?)  We're told that it's a online poll using "quality ‘research only’ panels." and that it is weighted, but not on what basis (beyond state).   I understand it is not intended to be a regular polling series.  

There are a lot of demographic breakdowns of primary vote gap - mostly unsurprising (younger voters are more pro-ALP but also more likely to be uncommitted to their vote, which is not the same thing as "undecided", etc).  These are of somewhat limited use because one needs to know the Green and minor party mix to really say how left-or-righr-leaning a given group may be; primary vote gap does not alone reveal it.  We're also told (p 4) that One Nation got 2% last election and indepedents got 8%; neither of these things are correct (try 4.96% and 5.29% respectively).  There's also an issues mix question which in common with any other poll this term finds cost of living way on top as the leading concern for voters.  

The more professional polling the merrier (though preferably with more weighting etc details than this) but the innumeracy, incompetence and bias of much of the mainstream media coverage of polling continues to do Australian voters a disservice.  There has been little if any improvement across the media in the standard of coverage since the 2019 polling failure and it seems that tabloid outlets in particular view polling as somewhere between a free story for their journalists and a launching pad for partisan spin and ragefarming about hung parliaments.  When these outlets do report polling exclusives the public cannot take any details on trust because these media just can't understand what is under their own noses.  I cannot see why such incompetence should be allowed.  

With a clear note that this award goes to the media outlets involved and not the pollster, the Murdoch tabloids and the Daily Fail share one of these for their hopeless coverage of this poll:

Porcupine Fish Award For Ultra-Fishy Poll Reporting (credit)