Federal Politics 2019-2022 – Page 37

Newspoll state leadership polling and Essential Research coronavirus latest

State-level polling finds the coronavirus tide lifting all boats — but none so far as Mark McGowan in WA, whose numbers may be without precedent.

The Australian ($) today provides Newspoll findings on state leaders’ handling of the coronavirus, from samples of around 520 for each mainland state plus 309 for Tasmania. The poll finds all concerned riding high, including three who strongly outperformed Scott Morrison’s ballyhooed 68% approval and 28% disapproval on the weekend. These are WA Labor Premier Mark McGowan, at 89% approval and 6% disapproval; Tasmanian Liberal Premier Peter Gutwein, at 84% approval and 11% disapproval after three months in the job; and Victorian Labor Premier Daniel Andrews, at 75% approval and 17% disapproval.

Morrison was also matched on approval and bettered on net approval by NSW Liberal Premier Gladys Berejiklian (69% approval and 23% disapproval) and SA Liberal Premier Steven Marshall (68% approval and 21% disapproval). Only Queensland Labor Premier Annastacia Palaszczuk, who faces an election in October, was below the prime ministerial par (55% approval and 39% disapproval). With due allowance for small samples, I believe McGowan’s ratings may be a record for Newspoll, or indeed for any other Australian pollster, and that Gutwein’s might have been too if not for McGowan’s.

The leaders record even stronger ratings on the specific question of handling the coronavirus outbreak: 77% rate Berejiklian as having done well, compared with 18% for badly; Andrews is at 85% and 11%; Palaszczuk is at 72% and 23%; McGowan is at 94% and 4%; Marshall is at 82% and 11%; and Gutwein is at 89% and 8%. Equivalent results are also provided for the Prime Minister, and here too Western Australians are most positive, at 73% approval and 23% disapproval, with 85% rating Morrison had handled coronavirus well compared with 14% for badly. In New South Wales, Morrison scored 67% approval and 30% disapproval, and 82% well and 16% badly for coronavirus; in Victoria, 72% approval and 26% disapproval, 83% well and 14% badly; in Queensland, 67% approval and 28% disapproval, 81% well and 17% badly; in South Australia, 70% approval and 27% disapproval, 83% well and 15% badly; and in Tasmania, 64% approval and 31% disapproval, 81% well and 18% badly.

As reported in The Guardian, the weekly Essential Research coronavirus poll provides us with a third set of small-sample findings on mainland state governments’ handling of the crisis, ranging from about 80 respondents in South Australia to 320 in New South Wales. The latest results produce combined very good and good ratings of 77% for the Victorian and South Australian governments, 76% for Western Australia, 67% for Queensland and 63% for New South Wales. The table below records the progress of this series over its three weeks, together with an averaged result which again shows Western Australia highest at 77%, followed by 74% for Victoria, 72% for South Australia, 61% for Queensland and 60% for New South Wales.

Essential Research also finds confidence in the federal government’s handling of the crisis continuing to rise, with 70% rating it good or very good, a measure that earlier progressed from 45% in late March to 65% last week. Seventy-three per cent now say they consider themselves unlikely to catch the virus, compared with 57% at the peak of concern at the end of March. In response to a list of options for budget repair, 64% supported preventing companies in offshore tax havens from receiving goverment support, but only 32% favoured removing franking credits and negative gearing, and 18% supported death duties.

On the COVIDSafe app, the weekend’s Newspoll found 21% saying they would definitely take it up, 33% that they would probably do so, 21% that they would probably not, and 18% that they would definitely not. Apart from the lower uncommitted rating, this is broadly in line with an Australia Institute poll of 1011 respondents on Thursday and Friday which had 45% saying they would and 28% that they wouldn’t. Essential Research also weighed in on the question, and found 53% saying it would limit the spread of the virus, and 46% that it would speed removal of distancing restrictions. A full set of results from Essential Research should be with us later today.

UPDATE: Full Essential Research report here.

Newspoll: 50-50

Scott Morrison gains further momentum in his remarkable but apparently voteless approval rating turnaround.

Courtesy of The Australian, the first Newspoll in three weeks is consistent with the last in suggesting the coronavirus surge in approval for Scott Morrison in translating into only a modest dividend on voting intention, on which the two parties are now tied after the Coalition opened up a 51-49 lead last time. On the primary vote, the Coalition is down a point to 41%, Labor up two to 36%, the Greens down one to 12% and One Nation down one to 4%. Despite that, Scott Morrison has gained further on his huge approval rating boost in the last poll, up seven to 68% — a level not seen since Kevin Rudd reached 70% in late 2008 — while his disapproval rating is down seven to 28%. Anthony Albanese is respectively steady at 45% and down two to 34% (I assume — the report says 36%, but this would be unchanged on last time), and Morrison’s lead as preferred prime minister is now 56-28, out from 53-29. The poll was conducted Wednesday to Saturday from a sample of 1519.

It’s as easy as APC

A new polling industry standards council takes shape; and the coronavirus polling glut keeps piling higher.

A promised initiative to restore confidence in opinion polling has came to fruition with the establishment of the Australian Polling Council, a joint endeavour of YouGov, Essential Research and uComms. Following the example of the British Polling Council and the National Council for Published Polls in the United States, the body promises to “ensure standards of disclosure”, “encourage the highest professional standards in public opinion polling” and “inform media and the public about best practice in the conduct and reporting of polls”.

The most important of these points relates to disclosure, particularly of how demographic weightings were used to turn raw figures into a published result. The British Polling Council requires that its members publish “computer tables showing the exact questions asked in the order they were asked, all response codes and the weighted and unweighted bases for all demographics and other data that has been published”. We’ll see if its Australian counterpart to sees things the same way when it releases its requirements for disclosures, which is promised “before July 2020”.

Elsewhere:

• The West Australian has had two further local polls on coronavirus from Painted Dog Research, one from last week and one from this week ($). The McGowan government announced its decision to reopen schools next week in between the two polls, which had the support of 22.7% in the earlier poll and 49% this week, with opposition down from 43.3% to 27%, and the undecided down from 34% to 24%. The earlier poll found remarkably strong results for the McGowan government’s handling of the crisis, with 90.0% agreeing it had been doing a good job (including 54.2% strongly agreeing) and only 2.9% disagreeing (1.2% strongly), with 7.1% neither agreeing or disagreeing. No field work dates provided, but the latest poll has a sample of 831.

• The University of Melbourne’s Melbourne Institute conducted a 1200-sample survey on coronavirus from April 6 to 11, and while the published release isn’t giving too much away, we told that “about 60% of Australians report being moderately to very satisfied with government economic policies to support jobs and keep people at work”, and that “more than 80% expect the impact of the coronavirus pandemic to last for more than 6 months“.

• The Washington Post’s Monkey Cage political science blog examines local government elections held in France on March 15, two days before the country went into lockdown: turnout fell from 63% to 45%, but the result was not radically different from the last such elections in 2016. Traditional conservative and socialist parties holding up well and the greens making gains, Emmanuel Macron’s presidential vehicle La République En Marche failing to achieve much cross-over success, and Marine Le Pen’s Rassemblement National losing ground compared with a strong result in 2014.

Essential Research coronavirus latest

Confidence in the federal government and other institutions on the rise, but state governments in New South Wales and Queensland appear to lag behind Victoria, Western Australia and South Australia.

The Guardian reports Essential Research’s latest weekly reading of concern about coronavirus finds satisfaction with the government’s handling of the crisis up two points to 65%, its best result yet out of the five such polls that have been published (no sign yet of the poor rating, which hit a new low of 17% – the full report later today should reveal all).

Last week’s question on state governments’ responses was repeated this week, and with due regard to sample sizes that run no higher than around 320 (and not even in triple figures in the case of South Australia), the good ratings have been 56% last week and 61% for New South Wales; 76% and 70% for Victoria; 52% and 63% for Queensland; 79% and 77% for Western Australia; and 72% and 66% for South Australia. Combining the results gives New South Wales 58.5% and Victoria 73% with error margins of about 3.7%; Queensland 57.5% from 4.6%; Western Australia 78% from 5.5%; and South Australia 69% from 6.9%.

Also included are Essential’s occasion question on trust in various institutions, which suggests that all of the above might be benefiting from a secular effect that has federal parliament up from 35% to 53% and the ABC up from 51% to 58%. The effect is more modest for the Australian Federal Police, up two points to 68%. In other coronavirus-related findings, the poll finds “half of all voters think it’s too soon to even consider easing restrictions“, with a further 14% saying they are prepared to wait until the end of May; that 38% said they would download the virus-tracing app, with 63% saying they had security concerns and 35% being confident the data would not be misused.

UPDATE: Full report here.

What the papers say

Random notes on coronavirus and opinion poll response rates, election postponements and a call to give counting of pre-poll votes a head start on election night.

No Newspoll this week it seems – which is unfortunate, because a report in New York Times ($) suggests coronavirus lockdowns are doing wonders for opinion poll response rates:

Even in online surveys, pollsters have also seen an increase in participation over the past few weeks. At the Pew Research Center, which does most of its polling through the online American Trends Panel, many respondents filled in a voluntary-comments box in a recent survey with expressions of gratitude.

It is inferred that “a wider variety of people are willing to tell pollsters what they think, so it’s more likely that a poll’s respondents will come closer to reflecting the makeup of the general population.”

Coronavirus is rather less conducive to the staging of actual elections, the latest casualty being the May 30 date that was set for Tasmania’s Legislative Council seats of Huon Rosevears, which was itself a postponement from the traditional first Saturday of the month. The government has now invoked a recently legislated power to set the date for a yet-to-be-determined Saturday in June, July and August. The Tasmanian Electoral Commission has expressed the view that a fully postal election, as some were advocating, did not count as an election under the state’s existing Electoral Act.

Tasmania and other jurisdictions with elections looming on their calendars might perhaps look to South Korea, which proceeded with its legislative elections on Wednesday. As reported in The Economist ($):

All voters will have their temperature taken before entering their polling station (those found to have fever or other symptoms will be directed to a separate polling booth). They will also have to wear a face mask, sanitise their hands and put on vinyl gloves before picking up a ballot paper and entering the booth. Election stewards will ensure people keep away from each other while queueing and voting. Door knobs, pencils and ballot boxes will be sterilised often.

Other than that, I can offer the following in the way of recommended reading: Antony Green’s post calling for pre-poll votes to be counted under wraps on election day starting from 2pm. This would address issues arising from the huge imbalance between election day booths, only one of which processed more than 4000 votes at the May 2019 federal election, and the three weeks’ accumulation of votes cast at pre-poll booths, of which 901 cleared 4000 votes, including 208 that went above 10,000 and ten with more than 20,000 (UPDATE: Make that 370 of more than 4000 and 208 of more than 10,000 – turns out the numbers in the table are cumulative). The result is that the largest pre-poll booths are not reporting until very late at night, many hours after the last trickles of election booths runs dry.

This has sometimes caused election counts to take on different complexions at the end of the evening — to some extent at the Victorian state election in November 2018, which ended a little less catastrophic for the Liberals than the election day results suggested, and certainly at the Wentworth by-election the previous month, when Liberal candidate Dave Sharma briefly rose from the dead in his struggle with the ultimately victorious Kerryn Phelps. It is noted that pre-poll votes in New Zealand are counted throughout election day itself, which is made practical by a ban on any election campaigning on the day itself, freeing up party volunteers for scrutineering who in Australia would be staffing polling booths.

Antony also argues against reducing the pre-poll period from three weeks to two, for which there has been quite a broad push since last year’s election, as it will lead to greater demand for the less secure option of postal voting, stimulated by the efforts of the political parties.

Also note my extensive post below on recent events in Wisconsin – you are encouraged to use that thread if you have something to offer specifically on American politics.

Ipsos Issues Monitor and much else

An Ipsos issue salience poll finds environmental concerns elbowed aside by health care and the economy. Also: prospects for elections amid coronavirus in Queensland and the Northern Territory, finalisation of the Brisbane City Council election, and a Greens plebiscite on leader selection.

Ipsos has published its latest Issues Monitor survey results, in which respondents are presented with 19 issues and asked to identify the three of greatest concern. This is conducted monthly but published haphazardly: the last set of results was published on the precipice of our new era back in January, but the accompanying chart in the latest release allows you to at least eyeball the results from February and March. The big news in January was an unprecedented concern about the environment in the wake of the bushfires, pushing that issue to the top of the pile for the first time. It has now been put into the shade by a massive surge of concern about three issues: health care, up since January from 31% to 55%; the economy, up from 25% to 47%; unemployment, for which the result went unpublished in January (it was at 18% in December) and is now at 39%. The environment is now equal fourth on 22%, down from 41%, where it ranks alongside cost of living, down from 31%.

The Ipsos release also features results on the best party to handle the five aforesaid issues, which are consistent with an improved standing for the Coalition since December, the last time comparable figures were published. Most striking is its improvement on its traditional weak spot of health care, on which it now leads Labor by 34% (up eight) to 29% (steady), with the caveat that 7% favoured the Greens and most of those would presumably have Labor as a second preference (a further 3% favoured another party). The Coalition has also widened its lead on the economy, up seven to 43% with Labor down to 22%, the Greens on 5% and others on 3%, and gained five on cost of living to 34%, with Labor steady on 29%, the Greens on 7% and others on 4%. It also holds a lead over Labor of 34% to 28% on unemployment, with the Greens on 6% and others on 3%, no figures from December being available for comparison in this case. On the environment, the Coalition is up four to 23%, Labor is down one to 17%, the Greens are on 28% and others are on 6%.

The poll was conducted online from a sample of 1000, and was presumably conducted over the previous fortnight, but all we are told is that the numbers are for April.

Further news:

• The count for the Brisbane City Council election is now all but complete, confirming a repeat of the 2016 result with the Liberal National Party winning 19 seats to Labor’s five and the Greens’ one, plus independent Nicole Johnston in Tennyson. The Greens came within an ace of taking Paddington off the LNP but no cigar, with LNP incumbent Peter Matic prevailing by 11,064 (50.7%) to 10,753 (49.3%) after preferences, a margin of 311 votes. Unlike at state elections, optional preferential voting prevails at Queensland’s local government elections, in this case to the advantage of the LNP since many preferences that might otherwise have flowed between Labor and the Greens instead exhausted. As Antony Green notes on Twitter, this certainly made the difference in Paddington, and might have saved the LNP from Labor in their three most marginal wards. LNP incumbent Adrian Schrinner’s winning margin over Labor’s Pat Condren for the Brisbane lord mayoralty was 286,297 (56.4%) to 221,309 (43.6%).

• Annastacia Palaszczuk has raised the prospect of Queensland’s October 31 state election being held entirely by post, as has been advocated by Opposition Leader Deb Frecklington. The ABC reports the next update of federal government modelling on the spread of the virus in around two weeks will guide a decision that is “expected in the next month”. However, it appears the Northern Territory election will proceed in a more-or-less normal fashion on August 22, with the Northern Territory News ($) reporting the local electoral commission does not consider a postal election an option because “postal services were still limited in many remote communities”. It is accordingly “looking to accommodate social distancing at the upcoming election by opening up more early voting centres, encouraging postal voting and extending early voting times”.

• The Greens are in the process of conducting a plebiscite of party members on how it will choose its leaders in future, the three options being the status quo of election by the party room, the “one member one vote” model of a straight ballot of party members, and Labor’s approach of a 50-50 model in which the result is evenly split between the two approaches. The party’s three former leaders, Bob Brown, Christine Milne and Richard Di Natale, jointly wrote a column in The Guardian that beseeched members to reject “one member one vote”, noting the disastrous consequences the model eventually had for the Australian Democrats, and gently suggested the status quo was to be preferred. A counter-argument was subsequently advanced in the same publication by five party luminaries, including current Senator Mehreen Faruqi and former Senator Scott Ludlam.

Essential Research and Morgan: more coronavirus polling

Two new polls suggest support for the federal government’s handling of the coronavirus crisis is still on the rise.

I’ll be taking part in the Political Geekfest videocast through Zoom with Peter Lewis of Essential Research and Katharine Murphy of the Guardian Australia at 1pm AEST today, which you can register for here. The subject of discussion will be this:

• The Guardian reports on another Essential Research poll focusing mostly on coronavirus, which would appear to be a weekly thing at least for the time being. The latest poll finds 59% rating the government’s response as about right, up from 46% last week and 39% in the two previous weekly polls; 13% rating it an overreaction, continuing its downward trajectory from 33% to 18% to 17%; and 29% rating it an underreaction, which bounced around over the first three weeks from 28% to 43% to 37%. Respondents were also asked to rate their state governments’ reactions, though with sample sizes too small to be of that much use at the individual level: the combined responses for very good and quite good were at 56% for New South Wales, 76% for Victoria, 52% for Queensland, 79% for Western Australia and 72% for South Australia. The poll also records a surprisingly high level of general morale, producing an average 6.7 rating on a scale of one to ten, unchanged from May last year. The full report should be published later today. UPDATE: Full report here.

• Also apparently a weekly thing is Roy Morgan’s coronavirus polling, which is being conducted online and not by SMS as I previously assumed – indeed, I believe this is the first online polling Morgan has ever published. Last week’s tranche showed a sharp rise in approval of the government’s handling of the matter from a week previous, with 21% strongly agreeing the government was handling the matter well (up twelve), 44% less strongly agreeing (up ten), 23% disagreeing (down ten) and 6% strongly disagreeing (down ten). Respondents had also become more optimistic since the previous week (59% saying the worst was yet to come, down 26 points, 33% saying the situation would remain the same, up 22 points, and 8% expecting things to improve, up four), and, contra Essential, slightly more inclined to consider the threat was being exaggerated (up five points to 20%, with disagreement down six to 75%). The poll was conducted last weekend from a sample of 987.

Of plagues and houses

Results finalised on Queensland’s two status quo state by-election results, and COVID-19 question marks over looming elections in New Zealand, the Northern Territory and for two Tasmanian upper house seats.

Counting has concluded for the Currumbin and Bundamba by-elections of a fortnight ago, with Laura Gerber retaining Currumbin for the Liberal National Party by a 1.5% margin against a 1.8% swing to Labor, and Lance McCallum retaining Bundamba for Labor by a 9.6% margin ahead of second-placed One Nation (UPDATE: Make that a 1.2% margin in Currumbin and 9.8% in Bundamba). As noted previously, the flow of Greens preferences to Labor in Currumbin was relatively weak, though not quite decisively so. Deep within the innards of the ECQ’s media feed, it says that Greens preferences were going 1738 to Labor (72.8%) and 651 (27.2%), though this can’t be based on the final figures since the Greens received 2527 rather than 2389 votes. Had Labor received 79.17% of Greens preferences, as they did in the corresponding federal seat of McPherson last May, the margin would have been pared back from 567 (1.5%) to 215 (0.5%).

I have three tables to illustrate the results in light of the highly unusual circumstances of the election, the first of which updates one that appeared in an early post, recording the extent to which voters in the two seats changed their behaviour with respect to how they voted. Election day voting obviously fell dramatically, as voters switched to pre-poll voting and, to only a slightly lesser extent, outright abstention. What was not seen was a dramatic increase in postal voting, which will require investigation given the considerable anecdotal evidence that many who applied for postal votes did not receive their ballots on time — an even more contentious matter in relation to the mess that unfolded in Wisconsin on Tuesday, on which I may have more to say at a later time.

The next two tables divide the votes into four types, polling places, early voting, postal and others, and record the parties’ vote shares and swings compared with 2017, the latter shown in italics. In both Currumbin and Bundamba, Labor achieved their weakest results in swing terms on polling day votes, suggesting Labor voters made the move from election day to pre-poll voting in particularly large numbers, cancelling out what had previously been an advantage to the LNP in pre-poll voting. This is matched by a particularly strong swing against the LNP on pre-polls in Currumbin, but the effect is not discernible in Bundamba, probably because the picture was confused by the party running third and a chunk of its vote being lost to One Nation, who did not contest last time.

In other COVID-19 disruption news:

• The Northern Territory government has rejected calls from what is now the territory’s official opposition, Terry Mills’ Territory Alliance party (UPDATE: Turns out I misheard here – the Country Liberal Party remains the opposition, as Bird of Paradox notes in comments), to postpone the August 22 election. Of the practicalities involved in holding the election under a regime of social distancing rules, which the government insists will be in place for at least six months, Deputy Chief Minister Nicole Manison offers only that “the Electoral Commission is looking at the very important questions of how we make sure that in the environment of COVID-19 that we do this safely”.

• After an initial postponement from May 2 to May 30, the Tasmanian government has further deferred the periodic elections for the Legislative Council seats of Huon and Rosevear, promising only that they will be held by the time the chamber sits on August 25. Three MLCs have written to the Premier requesting that the elections either be held by post or for the terms of the existing members, which will otherwise expire, to be extended through to revised polling date.

• The junior partner in New Zealand’s ruling coalition, Winston Peters of New Zealand First, is calling for the country’s September 19 election to be postponed to November 21, which has also elicited positive noises from the opposition National Party. It might well be thought an element of self-interest is at work here, with Peters wishing to put distance between the election and a donations scandal that has bedeviled his party, and National anticipating a short-term surge in government support amid the coronavirus crisis. Prime Minister Jacinda Ardern may be softening in her opposition to the notion, saying earlier this week it would “depend on what alert level we are at”. There has regrettably been no polling of voting intention in New Zealand in two months, although the government recorded enormously encouraging results in a Colmar Brunton poll on handling of the pandemic in New Zealand and eight other countries, conducted last week.