Turnbull loses 20th successive Newspoll, 54-46, but Yes to SSM support falls
- Written by Adrian Beaumont, Honorary Associate, School of Mathematics and Statistics, University of Melbourne
This week’s Newspoll, conducted 21-24 September from a sample of 1700, gave Labor a 54-46 lead, a one point gain for Labor since the last Newspoll, three weeks ago. Primary votes were 38% Labor (steady), 36% Coalition (down 1), 9% Greens (steady) and 8% One Nation (steady). This Newspoll is Turnbull’s 20th successive Newspoll loss; Abbott had 30 in a row before he was dumped.
37% (up 2) were satisfied with Turnbull’s performance, and 52% (down 1) were dissatisfied, for a net approval of -17. Shorten’s net approval was unchanged at -20. Turnbull’s lead as better PM was reduced from a blowout 46-29 three weeks ago to a more normal 42-31.
Turnbull’s ratings improvement is probably due to right-wing voters’ approval of his pro-coal policy. However, as I suggested here, this policy does not appear to be helping the Coalition in voting intentions.
Turnbull’s continued lead over Shorten as better PM is partly because Greens voters are uninspired by Shorten, and select “don’t know” when asked who is the better PM. In the most recent poll to give a better PM breakdown by voting intentions, the Greens backed Shorten by just 42-25 in Essential last fortnight.
SSM plebiscite polling
The same sex marriage plebiscite is currently in progress. Voters have until 7 November for ballot papers to be received by the ABS. The result will be declared on 15 November.
In Newspoll, 67% said they would definitely vote and 15% said they had already voted, so 82% would definitely or had already voted, up from 67% in August. Among those who would definitely or had already voted, Yes led by 61-34 (67-31 in August). For the whole sample, Yes led by 57-34 (63-30 in August).
In this week’s Essential, Yes led by 72-26 among the 36% who had already voted, and by 57-39 among the 45% who will definitely vote. Last week, Yes led by 63-33 among the 62% who would definitely vote, and by 59-37 among the 9% who had already voted. 81% this week say they will definitely or have already voted, up from 71% last week and 62% three weeks ago. Definite voters were 68-28 Yes three weeks ago. For the whole sample, Yes led 58-33 (55-34 last week, 59-31 three weeks ago).
Last week’s YouGov had Yes leading by 59-33, unchanged from a month ago. There were no breakdowns by likelihood to vote.
A potential problem with this polling is that pollsters are asking whether voters support changing the law, not how they will vote on the actual survey. There could be people who support changing the law, but will vote No because they want religious freedoms guaranteed, or are concerned about safe schools or other No campaign issues. So far Newspoll is the only pollster asking the correct question.
There is a large difference between Newspoll and Essential on the percent who have already voted (15% in Newspoll, 36% in Essential), even though these polls had similar fieldwork dates. I think Essential is more likely to be correct, as those who are keen to vote will do it shortly after receiving their survey.
I believe that the Yes lead has been narrowing somewhat because the Yes case is associated with the left, and conservative voters are wary of voting for anything that the left supports. However, Yes still has a big lead, and should win easily. If comparisons are made to Trump or Brexit, neither trailed by over 20 points, although UK Labour did face such a deficit before a huge surge saw the Conservatives lose their majority at the 2017 UK election.
In other polling on the plebiscite, Newspoll found voters opposed the postal plebiscite 46-44, a reversal of a 49-43 favourable result in August. Voters favoured providing religious guarantees by an unchanged 62-18. However, only 20% in Essential were very concerned about the impact of same sex marriage on religious freedoms, with another 15% concerned.
Kevin Bonham has a long article about the polling for the same sex marriage plebiscite.
ReachTEL polls in coal country seats show support for renewables
The left-wing Australia Institute commissioned ReachTEL to conduct polls in the NSW Federal seats of Hunter and Shortland on 15-16 September; these seats are both in the Hunter valley region. The sample was 643 in Shortland and 714 in Hunter. Labor held a 60-40 lead in Hunter and 58-42 in Shortland by respondent allocated preferences, with both seats swinging to the Coalition by about 2 points since the 2016 election.
The Hunter region is well-known for coal, and the Liddell coal-fired power plant is located here. However in both seats, more people supported AGL’s decision to close the Liddell plant than were opposed, and more thought energy from renewables was cheaper to produce than from coal. By margins over 20 points, people would prefer investment in renewables to coal.
Essential 53-47 to Labor
In this week’s Essential, Labor led by 53-47, a one point gain for Labor since last week, but a one point gain for the Coalition since last fortnight. Essential uses a two-week rolling average for its voting intentions polling (sample 1800). I believe Labor had a strong result three weeks ago, which was replaced by a weak result last week, but a stronger poll this week. Primary votes were 37% Labor (up 1 since last week), 37% Coalition (down 1), 10% Greens (steady), 7% One Nation (down 1) and 3% Nick Xenophon Team (steady).
Belief or disbelief in various propositions to do with aliens and the supernatural were ascertained, along with some conspiracy theories. By 68-21, voters did not believe that global warming is a hoax. By 58-16, they did not believe that vibrations from wind farms cause long-term health damage. By 70-14, they did not believe that vaccines can cause autism.
Over 80% agreed with three statements saying the government should do more to restrict private health insurance fee increases. By 60-27, voters thought private health insurance not worth the money paid for it.
In last week’s Essential, 64% (up 4 since February) thought climate change was human-caused, while 24% (down 1) thought it may be due to normal fluctuations. Belief in human causes has been trending up since a 48-39 margin in October 2012. 56% (up 7 since December 2016) thought Australia is not doing enough to address climate change, 20% (down 2) thought we were doing enough, and 8% (down 3) thought we were doing too much.
Among workers, 52% said they had not had a wage increase in the last year, while 36% said they had.
YouGov primary votes: 35% Labor, 34% Coalition, 11% Greens, 9% One Nation
In last week’s YouGov, conducted 14-18 September from a sample of 1060, primary votes were 35% Labor (up 3 from three weeks ago), 34% Coalition (steady), 11% Greens (down 1), 9% One Nation (steady), 3% Nick Xenophon Team (down 1) and 3% Christian parties (steady).
Despite the improvement for Labor’s primary vote, the two party result was an unchanged 50-50 tie. The Poll Bludger says these primary votes would by 54-46 to Labor by last election preferences. Other polls have been 1-2 points worse for Labor when using respondent preferences as compared with the previous election method, but a four point difference is far too large to be credible.
Authors: Adrian Beaumont, Honorary Associate, School of Mathematics and Statistics, University of Melbourne