Month: April 2021

The latest fortnightly Essential Research poll includes the regular monthly leadership ratings, their principal item of interest in between their quarterly dumps of voting intention numbers. The pollster included a bonus result for Scott Morrison’s leadership ratings in the last fortnightly poll by way of discerning any emergent gender gap in light of recent events. The results chart a steady decline for Scott Morrison, from 62% a month ago to 57% a fortnight ago to 54% now, and a corresponding rise in disapproval from 29% to 35% to 37%. While he remains well in positive territory, a distinct downturn can be observed in the BludgerTrack polling trend. The gender gap that opened a fortnight ago, which you can read about here, has neither narrowed nor widened.

Anthony Albanese records his weakest personal ratings in a while, with approval down two to 39% and disapproval up two to 34%. GhostWhoVotes, who monitors these things, points out that breakdowns by voting intention have him down five on approval among Labor voters to 55% and up six on disapproval to 22% – this is from a sub-sample of 483 and a margin of error of about 4.5%, so make of it what you will. In any case, he has taken a reasonable bite out of Morrison’s lead as preferred prime minister, which narrows from 52-26 to 47-28.

The remainder of the survey is mostly about COVID-19 and the vaccine rollout, including the regular question on the quality of the federal and state governments’ responses. The federal government’s good rating is down eight points to 62% and its poor rating is up five to 17%, though this isn’t a whole lot different to the situation before the government’s numbers surged in mid-November for whatever reason. The ratings for the five mainland state governments are down as well, by two in the case of New South Wales to 73%, four for Victoria to 58%, three for Queensland to 72%, seven for Western Australia to 84% and ten for South Australia to 75% (with progressively increasing caution required for small sub-sample sizes). As with the federal results though, these numbers don’t look that remarkable when compared with their form over the longer term.

Respondents were also asked how confident they would have been about COVID-19 management “if a Labor government under Anthony Albanese had been in power”, with an uninspiring 44% rating themselves confident and 37% not so. Fifty-two per cent felt the vaccine rollout was proceeding too slowly, with 19% happy with the situation and 20% signing on to the seemingly odd proposition that it was happening too fast. For those in the former category, 42% held the federal government mostly responsible, 7% state governments, 24% international supply chains and 18% “unavoidable delays in the production of vaccines”.

There are a whole bunch of further questions on the vaccine rollout, interstate travel and the end of the JobKeeper and JobSeeker supplements, plus one on paid parental leave, which you can read about in the full release.

I had a vague hope that last fortnight’s sooner-than-expected Newspoll portended a return to a (usually) fortnightly rather than three-weekly schedule, but apparently not. Essential Research should be along this evening though, hopefully including its monthly leadership ratings. Then there’s this:

• Roy Morgan has published results of an SMS poll conducted on Friday and Saturday from a sample of 1423, which asked if respondents were still up for the COVID-19 vaccine in light of recent developments (only 17% were not, increasing to 24% when those who would only settle for Pfizer were included) and, most interestingly, if they approved of Scott Morrison’s handling of “all COVID-19 related issues”. In the absence of a non-response option, the latter question recorded 49% approval and 51% disapproval. I’m aware of two past polls that specifically asked about leaders’ rather than governments’ handling of COVID-19, both from Newspoll – one in April and one in July – from which the weakest result was 61% approval and 36% disapproval for Daniel Andrews in the July poll.

• Sarah Elks of The Australian reports that Andrew Laming has declined Scott Morrison’s invitation to sit out the preselection for his Brisbane seat of Bowman, and is “collecting evidence in an effort to disprove a series of allegations against him”. The report notes he has an incentive in a $105,600 reward available to to MPs who “retire involuntarily”, which would not be granted if he went gracefully. Laming will also need to pass muster with the Liberal National Party’s “candidate suitability panel” if his nomination is to proceed to the stage where local party members have a say.

• In a piece for The Conversation, Benjamin Reilly of the University of Western Australia evaluates the likely impact of optional preferential voting, as mooted by the Coalition members of the Joint Standing Committee on Electoral Matters. The conclusion is that Labor would only have won about half of the 36 seats it won on preferences at the 2013, 2016 and 2019 elections after trailing on the primary vote, and that few if any members of the House of Representatives front bench would have got their foot in the door.

• I have a guide up for the Upper Hunter state by-election in New South Wales on May 22, though it’s still at a preliminary stage since most of the candidates haven’t been announced, together with a Tasmanian state election guide that has lately been supplemented with a page for the Legislative Council contests.

My assertion in the previous post that we faced a dry spell on the polling front hadn’t reckoned on Newspoll’s quarterly breakdowns, published today in The Australian. These combine the four Newspoll surveys conducted this year into a super-poll featuring various breakdowns from credible sample sizes (though I’d note that nothing seems to have come of talk that new industry standards would require that such breakdowns be provided in each poll individually, in a new spirit of transparency following the great pollster failure of 2019).

The latest numbers offer some particularly interesting insights into where the Coalition has been losing support over recent months. Whereas things have been reasonably stable in New South Wales (now 50-50 after the Coalition led 51-49 in the last quarter of 2020) and Victoria (where Labor’s lead narrows from 55-45 to 53-47), there have been six-point shifts in Labor’s favour in Western Australia (where the Coalition’s 53-47 lead last time has been reversed) and South Australia (51-49 to the Coalition last time, 55-45 to Labor this time). Labor has also closed the gap in Queensland from 57-43 to 53-47.

It should be noted here that the small state sample sizes are relatively modest, at 628 for WA and 517 for SA, implying error margins of around 4%, compared with around 2.5% for the larger states. I also observed, back in the days when there was enough state-level data for such things to be observable, that state election blowouts had a way of feeding into federal polling over the short term, which may be a factor in the poll crediting Labor with a better result than it has managed at a federal election in WA since 1983.

The gender breakdowns notably fail to play to the script: Labor is credited with 51-49 leads among both men and women, which represents a four-point movement to Labor among men and no change among women. There is also nothing remarkable to note in Scott Morrison’s personal ratings, with deteriorations of 7% in his net rating among men and 8% among women.

Further results suggest the government has lost support more among the young (Labor’s lead is out from 61-39 to 64-36 among those aged 18 to 34, while the Coalition holds a steady 62-38 lead among those 65 and over), middle income earners (a three-point movement to Labor in the $50,000 to $100,000 cohort and four-point movement in $100,000 to $150,000, compared with no change for $50,000 and below and a two-point increase for the Coalition among those on $150,000 and over), non-English speakers (a four-point decline compared with one point for English speakers) and those with trade qualifications (a four-point movement compared with none among the university educated and one point among those without qualifications).

You can find the full results, at least on voting intention, in the poll data feature on BludgerTrack, where you can navigate your way through tabs for each of the breakdowns Newspoll provides for a full display of the results throughout the current term. Restoring a permanent link to all this through my sidebar is part of the ever-lengthening list of things I need to get around to.