Making sense of the polls

Making sense of the polls

High Court rules Joyce and Roberts ineligible. SSM plebiscite turnout high

In July and August, six Senators and Deputy PM Barnaby Joyce were referred to the High Court as they may have been ineligible to stand for election or sit in Parliament. These cases were all considered under Section 44(i) of the Constitution, relating to having a foreign citizenship.

Greens Senators Larissa Waters and Scott Ludlam had resigned from the Senate once their dual citizenship was discovered, but the High Court still had to determine if they were validly elected. Nationals Senators Fiona Nash and Matt Canavan, One Nation Senator Malcolm Roberts and Senator Nick Xenophon did not resign.

Today, the High Court ruled that Waters, Ludlam, Nash, Roberts and Joyce were all ineligible to have been elected at the 2016 election, while Canavan and Xenophon were eligible. The four Senators found ineligible will be replaced after a special recount by the next on their party’s ticket.

As Joyce sits in the lower house, a by-election in his seat of New England is required, and will be held on 2 December. Joyce has divested himself of his NZ citizenship, and will run for this by-election. Independent Tony Windsor, who Joyce defeated 58.5-41.5 at the 2016 election, will not contest the by-election, so Joyce should win easily.

With New England vacant, the Coalition still holds 75 of the 149 current House seats, a bare majority. Unless Joyce loses the by-election, the Coalition retains its majority.

There are often big swings at by-elections, but in most by-elections, the government’s majority is not at risk. The stakes are higher in this by-election, which should help Joyce to mitigate the swing against him. It is very likely that Joyce will retain New England, and return to Parliament.

SSM turnout data and polling

As at last Friday 20 October, the ABS estimated it had received 11.9 million forms for the same sex marriage plebiscite (74.5% of the electorate). The ABS’s estimate is now based on forms scanned, rather than on the weight of containers with forms. On 13 October, an estimated 10.8 million forms had been returned under the old method. Only 300,000 forms were returned in the following week, so the ABS believes the old method was an underestimate.

In this week’s Essential, 75% said they had already voted, and a further 8% said they would definitely vote, implying a turnout of 83%. Those who had already voted favoured Yes by 60-34, while the remaining 25% favoured Yes 39-33.

High turnout could help Yes win an absolute majority of the whole electorate, not just of those who vote. Such a victory would give Yes more legitimacy, making it harder for conservative politicians to excuse delaying parliamentary action.

Today is the nominal deadline to post the envelope, but envelopes will be accepted until 6pm on 7 November. The result will be declared on 15 November.

ReachTEL 53-47 to Labor

A Sky News ReachTEL poll, conducted 25 October from a sample of 2400, gave Labor a 53-47 lead, unchanged since late September. ReachTEL is using respondent allocated preferences. At this stage, media reports do not include the forced choice that would have been asked of the 9% “undecided”. From the limited information that has been released, Kevin Bonham estimates the two party vote by last election preferences was 53.7-46.3 to Labor.

Turnbull’s better PM lead over Shorten narrowed to 51-49 (52-48 in September). ReachTEL uses a forced choice question for better PM, which tends to favour opposition leaders more than polls with a don’t know option. Shorten is much more likely to win a ReachTEL better PM poll than Newspoll.

39% of respondents had an NBN connection, and by 44-35, these people were dissatisfied with their connection. 92% are concerned about electricity prices increasing in the next year, including 68.5% very concerned. 52% selected cutting power prices as the most important priority, with 27% for reducing emissions and 20% reliability. Just 28% had heard a lot about the National Energy Guarantee.

This poll was taken Wednesday night, before most of the public were aware that the AWU investigation had backfired on Michaelia Cash. While the Cash affair is embarrassing for the Coalition, it may not move the polls as the public is cynical about all politicians, and expects some bad behaviour.

Essential 52-48 to Labor

This week’s Essential, from a sample of 1860, is unchanged on last week, with primary votes of 37% Coalition, 36% Labor, 9% Greens, 8% One Nation and 3% Nick Xenophon Team. Last week’s sample appeared bad for Labor, and that sample is still in the two-week aggregate. Labor’s lead may increase when that sample washes out next week.

By 35-18, voters approved of the NEG, with 47% unsure. By 35-32, they approved of not having a Clean Energy Target. By 41-32, voters disapproved of phasing out renewable energy subsidies by 2020. 31% thought power prices would increase as a result of the NEG, 16% decrease and 31% thought it would make no difference.

Voters would trust Labor by zero to 12 points over the Liberals on a range of energy issues. On all issues, at least 50% thought there was no difference between the major parties or didn’t know.

ReachTEL seat polls: Kooyong, Warringah and Wentworth

ReachTEL conducted three seat polls on 19 October for the left-wing Australia Institute. In Josh Frydenberg’s Kooyong, the Liberals led by 57-43, a 6 point swing to Labor since the 2016 election. In Turnbull’s Wentworth, the Liberals led by 57-43, an 11 point swing to Labor. In Abbott’s Warringah, the Liberals led by 60-40, a mere one point swing to Labor. Samples were 850-920 for each seat.

I think the low swing in Warringah is because ReachTEL asked for parties, not candidate names; this is reasonable as the identities of Labor’s candidates are unknown. Abbott is likely to be a significant drag on the Liberal vote in Warringah should he re-contest at the next election.

Japan election: landslide for governing coalition

The conservative LDP, with its Komeito ally, has governed Japan since its foundation in 1955, with only two brief periods in opposition. An election for Japan’s lower house was held on 22 October.

In Japan, electors receive two votes, one for a single member electorate and one for a proportional block. Unlike NZ and Germany, there is no attempt to compensate parties that do badly in electorates using the proportional allocation; these seats are simply added to electorate seats. At this election, there were 465 seats (down 10 from 2014), with 289 electorate seats and 176 proportional block seats.

In the electorates, the LDP steamrolled the opposition, winning 218 of the 289 electorates, with a further 8 electorates for its Komeito ally. In the proportional blocks, the governing coalition won 87 of the 176 seats, with 49 for the centre-left coalition and 40 for the populist right Hope Party and its ally. The LDP benefited from a divided opposition in the electorates, winning 48.2%, with 20.6% for their nearest opponent, the Hope Party.

Overall, the governing coalition won 313 of the 465 seats, down 11 from 2014, but still more than a 2/3 majority, with the LDP alone winning 284 seats. The centre left won 69 seats, the populist right 61 and Independents 22.