Poor ReachTEL for Labor, but Essential better

This week’s ReachTEL was reported as showing a 50-50 tie, unchanged on last week. However, this was based on respondent allocated preferences, and the primary vote movements imply a 2 point gain for the Coalition by 2013 election preferences, to a 52-48 Coalition lead. Labor is not at 50-50 when trailing the Coalition 42.7-33.2 on primary votes with the Greens at 9.9%. This poll was conducted 9 June from a sample of 2180.

Turnbull’s (total good) minus (total poor) rating was -9, up 5 points from last week, but below his -7 rating last fortnight. Shorten’s (total good) minus (total poor) rating was also -9, up 2 points.

In Essential, Labor had a 51-49 lead by 2013 election preferences, up one point on last week. Turnbull’s net approval was -2, down four points from last fortnight and the same as his net approval four weeks ago. Shorten’s net approval was -6, up four points on last fortnight. Essential was conducted 2-5 & 9-12 June from a sample of 1790, with questions other than voting intentions based on only last week’s sample.

Turnbull’s better PM lead in ReachTEL was 9 points, unchanged on last week. In Essential, he led by 11 points, down from 13 two weeks ago.

I would like to see more polls, but ReachTEL is a credible poll, and that result implies that the Coalition has pulled ahead on two party preferred (2PP), after a long period of ties.

Kevin Bonham’s poll aggregate is now at 50.3% 2PP to the Coalition, a 0.5 point gain for the Coalition since last week. The Poll Bludger’s BludgerTrack is now at 50.5% 2PP to the Coalition, a 0.7 point gain for the Coalition. Primary votes are 40.8% for the Coalition, 33.4% for Labor, 10.7% for the Greens and 4.2% for the Nick Xenophon Team (NXT). Since last week, Labor’s primary vote has dropped 1.1 points.

There have been some individual seat ReachTEL polls taken for Fairfax Media and 7 News. These suggest that Labor is not winning the swing required in the marginals. However, the NXT could do very well in SA, with polls implying they will win Mayo and Grey from the Coalition.

Notes on these polls

Consistent with poll movement to the Coalition, their position has improved in the last fortnight on all issues surveyed. The Coalition leads by 9 points on the economy, up from 5, and by 16 points on border protection, up from 12. Labor leads by 21 points on health, down from 25, and by 14 points on education, down from 21. Health and education spending were supported over company tax cuts by 70-30.

In last week’s Essential, 14% said they were taking a lot of interest in the election campaign, 39% some interest, 27% little interest and 14% no interest. 33% said their opinion of Turnbull had become less favourable over the last few weeks, with 7% more favourable. Shorten was perceived less favourably by 21% and more favourably by 20%.

Labor’s scores on a range of issues improved from May, particularly on the economy (Liberal by 20 to Liberal by 12). These questions are marred by high don’t know rates of at least 34% on all issues. The Liberals were seen as far better for corporate interests (Liberals by 41), with Labor ahead on most other economic issues. 45% disapproved of the business tax cuts, with 28% approving.

59% thought that climate change was caused by human activity, and 28% thought it was just a normal fluctuation in the earth’s climate; in March, this was 63-27 in favour of human activity. Most of the survey was taken before the recent NSW storms. Renewable energy incentives (49%) were the most supported potential action on climate change.

In this week’s Essential, Labor was up on all positive attributes since March, particularly on “has a good team of leaders” (up 9), good policies, vision for the future and “clear about what they stand for” (all up 8). The biggest change for the Liberals was a 9-point reduction in “divided”. The largest differences between the parties are in “too close to big corporate interests” (Liberals by 31), “out of touch with ordinary people” (Liberals by 17) and “looks after the interests of working people” (Labor by 25).

By 47-22, voters approved of superannuation changes made in the budget. 64% said they would vote on election day, and 23% before election day.

How the new Senate system will work

The new Senate system requires voters to number at least six squares above the line. This means that preferences will be decided by voters, and not by parties. There is a savings provision, so that voters who just vote “1” above the line will still be counted, but their votes will exhaust within their chosen party, and no preferences will be distributed to other parties from such votes.

Below the line voters are required to number at least 12 squares, but only six are needed for a formal vote.

Since this election is a double dissolution, 12 Senators will be elected in each state, rather than the normal 6. The quota for election is thus 1/13 of the vote, or 7.7%.

It is likely that the major parties and the Greens will fill 10 or 11 of the 12 quotas, in each state except SA, on primary votes alone. The last one or two spots are likely to be decided by partial quotas. A party’s remaining vote after electing all quotas filled on primary votes is important.

If the remainder is over 0.8 of a quota, that party will definitely win an additional seat, while a remainder of less than 0.2 of a quota will definitely not win more seats. Parties need a remainder of about 0.5 a quota to have a reasonable chance of winning extra seats.

Because the new system ended the group voting tickets that allowed micro parties to win from very small portions of a quota, micros will now also need at least 0.5 a quota to have a reasonable chance of election.

There will be more preferences than under fully optional preferential voting, which is used in NSW state elections. However, preferences are unlikely to alter the results unless the gap is less than 0.2 of a quota.

The new Senate system should have persuaded micro parties to merge, so as to concentrate their votes and maximise their chances of winning seats. This has happened to some extent on the left, with the Science Party and Cyclists Party running joint tickets in some states, and similarly with the Sex Party and Marijuana Party.

However, the right wing micros have not merged. For example, there are two Christian parties on the Victorian, Queensland and WA Senate ballot papers - Australian Christians and Fred Nile Christians. As there are so many right micros to choose from, it is doubtful that there will be enough votes concentrated with one right micro to win seats.

Other than the Coalition, Labor, the Greens and the NXT, I think Jacqui Lambie in Tasmania and Derryn Hinch in Victoria have the best chances of winning seats. Hinch benefited from drawing column A on the Victorian Senate ballot paper. Glenn Lazarus and Pauline Hanson in Queensland, and Family First’s Bob Day in SA, are also chances.

It is very unlikely that the Coalition will win a Senate majority. The best plausible outcome for the Coalition would be that the Coalition plus the NXT would have a majority, allowing the Coalition to pass legislation opposed by Labor and the Greens with only Xenophon’s support.

Antony Green has detailed analysis of the Senate in each state.

Impact of preference decisions

The Liberals' decision to put the Greens behind Labor on their How to Vote cards will greatly assist Labor in the seats they are defending from the Greens; this applies to Batman and Wills in Melbourne, and Grayndler and Sydney in Sydney. In these seats, the Greens will now need to have a substantial primary vote lead over Labor.

In SA, Labor and the Liberals will be issuing open tickets, leaving it up to voters whether to put the NXT ahead of the traditional enemy. In my opinion, the vast majority of Liberal and Labor voters will put the NXT ahead, even without How to Vote material.