Poll Roundup: No Instant Damage After Shocking Week

2PP Aggregate: 53.4 to Labor (unchanged)
Labor would win election "held now" comfortably

Last week the Coalition had an extremely messy six-hour party room meeting on the issue of same-sex marriage.  The meeting led to open infighting between Coalition MPs and a general perception of a shambles.  The end result was a decision (by a roughly two-to-one vote of the joint partyroom) that the Coalition would not allow a "free vote" on a cross-party same-sex marriage bill introduced by Warren Entsch.  (Backbenchers can still exercise a conscience vote, but any frontbencher who does can expect to join them there.)

At a Prime Ministerial press conference, Tony Abbott stated afterwards that it was his "strong disposition" that a plebiscite be held after the next election.  Malcolm Turnbull has since said that the party room has not yet made a specific decision to adopt the plebiscite as policy, and suggested it be held before the next election.  Scott Morrison has instead called for a referendum.

The week ended with the Coalition, still reeling from the dumping of Bronwyn Bishop, facing a fresh scandal over the neutrality or otherwise of Trade Unions Royal Commissioner Dyson Heydon, who accepted an invitation to speak at a Liberal Party event while serving in that role.  Combined with Heydon's attack on Bill Shorten as a witness, there is a perception about that Heydon is too politically biased, or at least too readily seen to be biased, to continue in his role.  The email trail has been dominating Question Time in the first days of this week.



The combination of the same-sex marriage mess, the Heydon saga, ongoing poor polling, reactions to the Coalition's climate target announcement and a general sense of lack of narrative led to a remarkable flood of damning media comment pieces. Declaring the government to be currently failing not just the sniff test and the pub test but also the intelligence test, a wide variety of writers (with the Fairfax presses to the fore) concluded that the government was just too stupid to survive.  Laura Tingle's piece comparing the government to the final year of Whitlam was especially acclaimed, though some thought the comparison harsh on Whitlam, who was at least always trying to actually do something.

The Coalition has been trying to flush the news cycle by any means available, be it the Prime Minister's sudden enthusiasm for more women in parliament, a scheme to dob in ice dealers, the PM's diet and exercise regime and so on.  Today we saw the Coalition take out its frustrations on green groups by seeking to amend environmental approvals law to overcome cases like the Adani coal mine court finding.  The funny thing there was that the Adani result happened because the government directly repeated a simple and easily avoidable procedural blunder made by the government before it.
Expectations were high that last week's shambles would deliver a massive poll whack for the Coalition, and many in the government would have been fearing the same.  We'll have to wait for next week to see the Newspoll response, but the polls out this week are not showing any action.  Whatever the subjective evidence, if we line up poll results for this government against others in the past, it's still far from clear this government is cactus.  Governments have been behind this long and won, have been further behind with less to go and won, have had PMs this much disliked and won - and have probably been widely condemned as stupid and won too.

This week's polls

The week could have done with more federal polls after all that, but as it was there were just two.  Fairfax-Ipsos came out with a 54:46 to Labor, a one-point move since last month.  Essential came out with only a 52:48, a one-point move back since last week.

After taking the primaries into account I aggregated the Ipsos at 54.1 to ALP and the Essential at just 51.8.  My aggregate weights Ipsos more heavily than Essential, so the two cancelled each other out and we're back  where we were, at 53.4.

The Ipsos result was greeted with glee by ALP supporters on social media, but I don't think that was justified.  A more reasoned response, after such a week, is really "is that all?"

Notably the Greens again polled very high in Ipsos (16% for the second month running).  There was also a two-point difference between Ipsos' last-election preference result and their respondent-preferencing result (56:44).  A rather striking chart from Poll Bludger shows that the Coalition's share of respondent-reported third-party preferences continues to shrink, with most recent polls giving the Coalition less than 30% of respondent-allocated preferences, and the latest Ipsos the lowest thus far at 22%.

At the last election the Coalition received about 38% of all non-major-party preferences.  Based on the Greens' increased share of the third-party vote it would be expected that this would now be around 33%.  The average for the last nine polls by various sources is around just 27% with a clear downwards trend. The impression in recent respondent polling is that a body of voters of whom two-thirds are Green and the remainder mostly voters for right-wing micros or independents, would preference Labor almost as strongly as the Greens did in 2010.  They might well say they'd vote that way, but I'd be greatly surprised if that occurred at an election.

Leaderships and Ministerial Approvals

Only Ipsos polled leadership results this week, with Tony Abbott down one to -24 (35-59) and Bill Shorten leading as preferred Prime Minister, 45-39, a two-point increase in his lead.  The big result was a ten-point jump in Shorten's netsat, up to -10 (39-49).  It will be interesting to see if other pollsters follow suit.  They have generally had Shorten starting to turn a corner, but nothing like this.  Two possibilities exist: firstly that Shorten is being looked at anew by voters in light of the new credence given to the idea that the Trade Unions Royal Commission is biased against him.  The second is that this is something specific to Ipsos, which took one month longer than other pollsters to register the previous plunge in Shorten's numbers.

We also saw new within-party preferred leader polling from Ipsos, and nothing too groundbreaking there.  Tony Abbott dropped back from 19% to 15% as preferred Coalition leader and still trails Turnbull (41) and Bishop (23) with Hockey and Morrison each scoring only 5.  On the Labor side, Bill Shorten has also lost ground since February, but still holds sway just, with 25% from Tanya Plibersek 23% and Anthony Albanese 19%.  Chris Bowen (8%) and Tony Burke (5%) round out the field.  Unlike the recent Morgan, these polls still show each current leader as the preferred leader among voters for that party, although 33% for Abbott and 34% for Shorten still aren't flash.

Essential had approvals for selected Coalition ministers.  These showed a massive rise since last year for Julie Bishop (now at +34) and Malcolm Turnbull also travelling well on +23.  Scott Morrison, who the commentariat continue to regard as a leadership contender but who never registers all that much with the voters, clocked in at -5.  Christopher Pyne, a much-ridiculed figure who is being reassessed following his performance on the same-sex marriage issue, rose to -9, but Treasurer Joe Hockey declined to -17.

Same-Sex Marriage Polling

In the week that the Coalition decided that just letting its MPs vote freely on same-sex marriage was all too hard, the new Ipsos recorded the highest level of support ever for a neutrally conducted poll (69-25; Crosby-Textor's 72-21 poll was commissioned).  Live phone polling is likely to overstate support by up to about eight points because some respondents are afraid to answer the question honestly in case the interviewer is gay.  Still, this is a very strong result, and one which showed that the constant attempts by opponents of marriage reform to confuse the debate with the red herring of same-sex parenting are failing here as they fail everywhere else.

One reason the Coalition's polling might not be as bad following last week's dustup is that those voters who support same-sex marriage but aren't deeply involved in the issue might see the outcome as a good one all considered.  Indeed, while I have argued here that a plebiscite is a dreadful way to resolve (or not resolve) the issue, it still beats the Coalition continuing to block reform until Labor wins an election by gagging its frontbench.

There are two polls on the plebiscite concept but the wording of one is badly flawed and the wording of the other is probably worse.

Canning Newspoll:

Firstly, Newspoll took the opportunity to ask voters in the seat of Canning the following question:

"The two options for deciding whether to allow same-sex marriage in Australia is [sic] a people's vote, where all people can vote in a referendum or plebiscite, or a vote by politicians in parliament.  On balance, do you believe the issue should be decided by a people's vote or the politicians in parliament?"  

Urgh.  For starters, it is not clear how a referendum can be a valid means of deciding the issue, unless it's a referendum to prevent the parliament adopting same-sex marriage, which no-one much seems to support [edit: a possible approach has been made clearer; see comments].  A plebiscite also doesn't bind the parliament; it would just be a giant opinion poll to add to all the ones we already have, with a decision from politicians to accept and implement it still required.  So the idea that a people's vote really decides the matter and eliminates politicians from the equation is a false dichotomy.  Also, the use of the term "politicians in parliament", as opposed to just "members of parliament" or "the parliament", would be likely to be heard pejoratively.

There are also problems with the idea that the issue would be resolved by a single vote by politicians.  More likely but for the plebiscite possibility, it would take more than one parliamentary vote for same-sex marriage to pass, after which no later vote would dare dismantle it.  There are also different shades of vote in parliament - for instance some voters might believe the issue should be decided by a free vote of MPs and not one where some MPs are gagged by their party.

The reading in Canning in response to this badly flawed question was 78% for the "people's vote" and 20% for the "politician's vote" with the response breaking 79-18 among Liberal supporters, 72-26 for ALP, and I estimate 85-14 from Others voters.  (Both Greens supporters and the Christian Right would have reasons to support a "people's vote" - Greens supporters because they fear the parliament might keep blocking but are confident the people wouldn't, the Christian Right because they fear the parliament would pass it but hope a popular vote wouldn't, or else at least cherish the platform they would get.)

Canning is likely to be reasonably representative nationally so I'd expect similar results elsewhere.  Question design would probably make some difference, but I doubt it would be enough to stop the basic response: that supportive voters would prefer a national vote to further Coalition obstruction.  Indeed, it's most likely that Coalition MPs who wish to hold up same-sex marriage or prevent it altogether, deliberately blocked a free vote so that they could make the parliamentary option seem worse than it is and thereby gain approval for a popular vote.

Marriage Alliance Sexton poll

Last month I commented at length on some extremely dodgy same-sex marriage polling (and even worse reporting) by Australian Marriage Forum.  Another poll commissioned by Australian Marriage Alliance (a different group) and administered by Sexton Marketing has been widely reported.  However at this stage full details of this poll are elusive.

One widely-reported finding has been that marriage equality ranked thirteenth on a list of priority issues for voters, although the full list hasn't been seen.  This is hardly unusual - relatively few voters see it as the most important issue, especially when asked to pick from a long list.

The more significant reported finding was 76% support for a plebiscite, similar to the Canning Newspoll result.  However, because the full list of questions has not been released, it is unclear whether the content of previous questions primed the respondent to answer in a certain way.  As priming is a routine feature of such anti-SSM polls as have been published, we should assume it was pending evidence otherwise.

What detail of the plebiscite question has been published suggests that voters were offered a choice between "politicians alone should vote on same-sex marriage" and "all Australians to first have a chance to have a say on this in a national vote".  The Australian's report (Leave gay marriage to us: voters) implies that the argument that the issue is "important" may have been used to prod responses towards the latter option.

There are also reports that respondents (49:44) said they "wanted to take time for an informed debate", but the context of that question is at this stage totally unclear.  Another finding reported by The Australian (Same sex marriage: Voters in no rush) is of 24% thinking same-sex marriage is a high priority and 35% medium.  However every single finding of this poll is unreliable without full publication of all details including the order of all questions.

I don't think it is difficult to word questions fairly on this issue.  A question like "Do you think a national plebiscite of all voters on same-sex marriage should be staged prior to Parliament next voting on the issue?" would cover it, and for the timing "If a plebiscite of all voters on same-sex marriage is to be held, should that be before, at or after the next federal election?" should be fine.

Other Polling

Both Ipsos (58%) and Essential (53%) found voters think "the government" and "Australia" respectively are not doing enough to address climate change.  32% and 35% respectively of Coalition voters agreed this was a problem.  Essential also found that non-Coalition voters think the government will try to "bring back industrial laws similar to Work Choices" (how you can bring back something that isn't quite the same escapes me, but I digress.) They also found that Coalition voters wouldn't much mind it if this happened, but other voters would.  Essential voters also considered government to be Australia's most corrupt of several sampled sectors, with science and technology the least.

I will have more on Canning in a separate article on that by-election soon.