Category: Federal Election Campaign 2016

Today’s stop in the six-part Senate tour is Queensland, for which the political and historical overview can be viewed here. Queensland was extraordinary in 2013 for the level of support for Palmer United, and it is thus notable for being the only state in which BludgerTrack has the “others” vote down from the election, by 5.9%. The Coalition is also down 2.7%, leaving room for Labor to increase 4.0% from their low base, and the Greens to improve 4.6%, after their vote fell by more than half in 2013. As in the previous posts, I will now set about attempting to project the result based on these swing figures, estimated preference flows based on how-to-vote cards and below-the-line preferences, and educated guesswork to fill a few gaps involving new parties, parties not recontesting, and the huge slump in support that awaits what’s left of Palmer United.

The big picture view of the count is that a clear two seats are available in the final stages for the remaining micro-parties. The Coalition starts the count with very near to five quotas, Labor with very near to four, and the Greens with barely more than one – the Coalition and Labor get their fifth and fourth candidates elected as the stragglers are eliminated from the count, the second Greens candidate (Andrew Bartlett, as it happens) drops out of contention at about the same time, and an array of seven micro-parties remain in the count with two seats left to spare. The three serious contenders out of the group are Glenn Lazaraus, Pauline Hanson and Katter’s Australian Party, who are respectively credited with base votes of 3.0%, 2.7% and 2.6%.

The Katter’s Australian Party total is determined by factoring the diminished projected share of the others vote on to their 2.9% total in 2013, but the scores for Lazarus and Hanson are purely based on guess work. The preference flows used for these candidates are also rather arbitrary, with a 20% bonus applied to the One Nation preference flow from 2013 in the case of Hanson, and Lazarus credited with three quarters of the flow to Palmer United. The KAP did particularly well on preferences in 2013, and on that basis its candidate is projected to pull well care of Lazarus and Hanson (both major parties are also directing a preference its way, although this ultimately doesn’t amount to much in the projection). That leaves Lazarus and Hanson battling for the final spot, with the projection calling it for Hanson by a tiny margin.

The safest way of interpreting this is that Queensland is set to return five Coalition, four Labor, one Greens and two micro-party Senators. The identity of the latter is up in the air, but there seems a strong chance that Hanson will be among them. Any candidate with a base vote approaching 3% will be very much in the game, and Hanson’s traditional stumbling block of major party preferences looks set to do her no harm at all, as the major parties will have only tiny surpluses to pass on to other candidates.

The Australian reports Newspoll shows the Coalition opening a 51-49 lead, from primary votes of Coalition 43% (up two), Labor 36% (steady) and Greens 9% (down one). Malcolm Turnbull is up one on approval to 37% and steady on disapproval at 51%, Bill Shorten is steady at 35% and down one to 50%, and Turnbull leads 45-30 as preferred prime minister. The poll of 1713 respondents was conducted Thursday to Sunday. Here’s the latest BludgerTrack update, including tonight’s Newspoll and yesterday’s Galaxy:

Here’s a closer look at how the minor party vote has tracked since the 2013 election, with the Greens shown in green, Palmer United in orange-brown, and others in grey.

The Sunday News Corp papers have a national federal poll from Galaxy, although their websites are being a little coy about the fact. The poll shows two-party preferred at 50-50, which is all I can tell you about voting intention at this stage, because I’m not seeing any primary votes, sample sizes or field work dates. (UPDATE: Primary votes here – Coalition 42%, Labor 35% and Greens 11%). The report does relate that a follow-up question found Labor would lead 53-47 if Tony Abbott was still Liberal leader; that 38% believe Labor’s claim that a Coalition government would privatise Medicare, compared with 45% who don’t; and that 30% believe Malcolm Turnbull’s claim that Labor’s negative gearing reforms would drive down house prices, compared with 40% who don’t. More to follow on that at a later time.

In other news, today’s Fairfax papers have a report canvassing party insiders’ views on the state of the horse race:

• A Nationals source is quoted saying the party is “pretty nervous” about Rob Oakeshott’s challenge to Luke Hartsuyker in Cowper, and “fearful of losing Page”, where Labor’s Janelle Saffin seeks to recover the seat she lost to Kevin Hogan in 2013. However, its polling is also said to show Barnaby Joyce leading Tony Windsor in New England.

• Labor is said to be confident about the outer Sydney seat of Macarthur, but less so about other Sydney marginals including Lindsay and Banks.

• In Central Queensland, Capricornia and Flynn are rated as “likely Labor gains”, while Nationals MP George Christensen is “precarious” in Dawson.

• In Victoria, Corangamite is said to be the only Liberal-held seat Labor is now targeting, suggesting it is not hopeful about the Melbourne seats of Dunkley, Deakin and La Trobe. The Labor-versus-Greens contest in Batman is rated as lineball, but Labor is thought unlikely to lose its vulnerable Melbourne seats of Chisholm and Bruce to the Liberals. Liberal candidate Chris Jermyn’s poor performance is thought likely to save Labor from the Country Fire Authority backlash in McEwen, but the controversy is giving the Liberals an “outside chance” in Bendigo.

Further:

• David Crowe of The Australian reports Jacqui Lambie is “performing so strongly in Tasmania that major party observers expect her to win and perhaps gain enough votes to elect her running mate, Devonport mayor Steve Martin”. The report also suggests the Nick Xenophon Team could potentially win seats in Victoria and Western Australia, and suggests Derryn Hinch, Bob Day and David Leyonhjelm are stronger prospects than Pauline Hanson and Glenn Lazarus, without writing either off (unlike Ricky Muir and John Madigan, who don’t rate a mention).

• Sarah Elks of The Australian reports on Labor efforts to shore up Terri Butler, its member for the inner southern Brisbane seat of Griffith, citing Liberal National Party insiders who say “the ALP has been panicked into throwing money at a seat it is no danger of losing”.

Welcome to part two in a hopefully six part series on the Senate contests. My political and recent historical overview of the Senate for Victoria can be found here – in this post, I shall stick to a one-shot attempt to model the result. The basis of the model and its assumptions are mostly as they were in the earlier New South Wales post. This time there is no complication of the mistaken Liberal Democrats vote to deal with, so the guide for determining vote shares is the swings currently recorded by BludgerTrack of -4.0% for the Coalition, -0.1% for Labor, +1.7% for the Greens and +6.1% for others.

The two immediately difficulties are in estimating the base votes for Derryn Hinch and the NXT and the flows of preferences they stand to receive, since the last election does not offer a guide. State-level polling suggests NXT is on around 2.5%, so for simplicity’s sake I’ve granted that much to both of them. The flow of preferences to Hinch is based on Palmer United’s, while the NXT gets what Xenophon received in South Australia at the 2013 election divided by three. I’m afraid I haven’t provided Ricky Muir with special treatment – no doubt he’ll do better than his 0.5% in 2013, but it’s my judgement that he won’t be seriously competitive and will be excluded in early rounds.

As shown in the table below, the result is five seats for the Coalition, four for Labor and two for the Greens, with the last seat to be won by a micro-party candidate – the Australian Sex Party on my projection, by the barest margin over Derryn Hinch. However, I strongly suspect I have underestimated Hinch’s vote, whereas I’d doubt I’ve gone too far wrong with the Australian Sex Party. The Sex Party, Hinch and NXT each start in reasonably close proximity on the primary vote, so the matter then comes down to the preference flows, which in the case of Hinch and especially the NXT are highly speculative. Hinch performs more strongly than NXT partly be virtue of receiving a preference recommendation from both major parties, but less strongly than the Sex Party, which did particularly well on below-the-line preferences in 2013 (which are used to estimate preference flows in the model), and also had a preference recommendation from Labor.

The availability of a seat for a micro-party is itself a close-run thing, as illustrated at Count 45, at which point the fifth Labor candidate is excluded in a very tight result against the NXT. Had Labor snuck its nose ahead at this point, the model projects their candidate ultimately winning the seat. This is based on the behaviour of Xenophon preferences in South Australia, of which the major parties were the biggest beneficiaries. Instead, the model finds the exclusion of Labor and the distribution of its preferences nudging the Sex Party clear of Hinch. It should be noted that Labor preferences had considerably more bearing on the result than Liberal preferences, since the fifth Liberal candidate had next to no surplus to distribute at the point of their election.

The latest weekly ReachTEL campaign poll for the Seven Network has two-party preferred unchanged at 51-49 in favour of the Coalition. However, the Coalition is down 1.1% on the primary vote to 42.4% on forced response primary votes, with Labor up 0.2% to the Greens up 1.3% to 10.5%, translating into a 1% shift to Labor if preference flows from the previous election are applied. The failure of this to translate into movement on the headline two-party result is down to a more conventional looking respondent-allocated preference result this week – and perhaps also to the fact that ReachTEL has dropped the Nick Xenophon Team from its list of options outside of South Australia, in recognition of the fact that it won’t be fielding lower house candidates anywhere else (correction – it does have a few candidates here and there). On personal ratings, Malcolm Turnbull records a tick upwards, from 27.4% to 33.5% on the combined very good plus good rating and from 36.3% to 33.3% on poor plus very poor, while Bill Shorten also improves, from 29.6% to 30.7% favourable and 39.7% to 37.8% unfavourable. Turnbull’s lead on preferred prime minister is slightly improved, from 57.6-42.4 to 58.4-41.6.

This week’s regular ReachTEL marginal seat campaign poll for Seven is from Cowper, and it provides more evidence of Rob Oakeshott being highly competitive in his bid to unseat Nationals MP Luke Hartsuyker. The primary votes are Nationals 42.2% (53.9% at the 2013 election post-redistribution), Rob Oakeshott 32.1%, Labor 11.1% (23.6% in 2013) and Greens 8.4% (10.9% in 2013). Based on a 72.7-27.3 respondent-allocated preference flow to Oakeshott, this translated into a two-party preferred result of 50-50.

We’ve also got marginal seat polling galore today courtesy of the News Corp tabloids, with Galaxy polling conducted for its Sydney, Melbourne, Brisbane and Adelaide papers, and ReachTEL going into the field for The Mercury in every seat in Tasmania. The Galaxy polls produce an average swing to Labor of around 2%, and are thus mostly disappointing for them, but the swing in the ReachTEL poll is closer to 3%, which in the Tasmanian context puts three seats on edge. Starting with the Galaxy polls, which surveyed slightly more than 500 respondents per electorate:

• The Daily Telegraph has polls of six Liberal-held marginals in New South Wales, showing every one going down to the wire, with the Liberals fortuitously poking their nose in front in every case but one. Two-party results are 52-48 in Banks (0.5% swing to Labor) and Reid (2.2% swing to Labor), 51-49 in Dobell (1.4% swing to Liberal), Gilmore (3.0% swing to Labor) and Lindsay (2.0% swing to Labor) and 50-50 in Macarthur (3.3% swing to Labor).

• The Herald Sun’s numbers suggest a status quo result across two Liberal-held and two Labor-held seats. The Liberals lead 53-47 in both Corangamite (0.9% swing to Labor) and Dunkley (2.6% swing to Labor), and Labor leads 52-48 in both Bruce (0.2% swing to Labor) and McEwen (1.8% swing to Labor).

• The Courier-Mail reports Labor leads of 54-46 in Petrie (4.5% swing to Labor) and 51-49 in Capricornia (1.8% swing to Labor), Liberal National Party leads of 52-48 in Brisbane (2.3% swing to Labor) and 53-47 in Longman (3.9% swing to Labor), and a 58-42 lead for Bob Katter in Kennedy (5.8% swing to Katter). Also polled was the Labor-held seat of Griffith, where Labor has reportedly been worried, but the poll records a 53-47 result in favour of Labor Terri Butler, unchanged on Kevin Rudd’s winning margin in the seat at the 2013 election.

• The Advertiser reports results of 50-50 in Hindmarsh (1.9% swing to Labor) and 53-47 to the Liberals in Boothby (4.1% swing to Labor). The Nick Xenophon Team was third in both seats, on 19% in Boothby and 16% in Hindmarsh.

ReachTEL’s Tasmanian polls bring better news for Labor, finding them leading in one of the three Liberal-held marginals and dead level in the other two. Denison and Franklin look set to remain with Andrew Wilkie and Labor’s Julie Collins respectively. The polls were conducted last night and have slightly smaller samples than we’ve been used to seeing from ReachTEL, presumably because Tasmania’s electorates themselves have only about three-quarters of those on the mainland. The results:

Bass (Liberal 4.0%): Nothing in it on two-party preferred, from forced preference primary votes of Liberal 42.6% (47.8% last election, 46.2% last poll) Labor 33.4% (34.6% last election, 36.0% last poll) and Greens 10.4% (7.9% last election, 9.7% last poll). The result on previous election preferences would be 51.2-48.8 in favour of Liberal. Sample: 538.

Braddon (Liberal 2.6%): Another tie on two-party preferred, from primary votes of Liberal 42.7% (46.9% last election, 46.4% last poll), Labor 37.9% (37.6% last election, 34.4% last poll) and Greens 8.8% (5.2% last election, 6.6% last poll). Labor has the edge on previous election preferences, at 51.0-49.0. Sample: 566.

Denison (Independent 8.9% versus Liberal): Andrew Wilkie has 34.5% of the primary vote (38.1% at the election, 37.3% last poll), the Liberals are second with 29.5% (23.2% last election, 27.3% last poll), Labor is third on 24.5% (24.8% last election, 22.1% last poll) and the Greens are on 8.7% (7.9% last election, 13.3% last poll). ReachTEL has a 63-35 two-candidate result for Wilkie versus the Labor candidate, but the final count would in fact be between Wilkie and the Liberal, not that it would make much difference to the result. Sample: 552.

Franklin (Labor 5.1%): Labor leads 59-41 from primary votes of Labor 37.1% (39.9% last election, 40.7% last poll), Liberal 37.6% (38.7% last election, 34.3 last poll) and Greens 18.3% (12.2% last election, 15.9% last poll). On previous election preferences, the result is 56.7-43.3. Sample: 550.

Lyons (Liberal 1.2%): Labor has a commanding lead of 55-45 in what has generally been reckoned its likeliest Tasmanian gain, from primary votes of Liberal 40.4% (44.4% last election, 45.8% last poll), Labor 35.2% (36.8% last election, 29.2% last poll) and Greens 11.8% (8.3% last election, 13.3% last poll). The result is a fair bit narrower on previous election preferences, at 51.1-48.9. Sample: 540.

Now here’s the latest BludgerTrack update, inclusive of the ReachTEL national result and (for state breakdown purposes) its Tasmanian polls:

The New South Wales Teachers Federation has produced the most intriguing set of marginal seat poll numbers of the campaign so far, showing Labor headed for victory in six crucial seats in New South Wales. The polls were conducted on Monday by ReachTEL, and target the same electorates as an earlier round of polling for the union on April 19. Five of the six results record movement to Labor since the earlier poll:

Dobell (Labor 0.2%): A poll of 616 respondents credits Labor’s Emma McBride with a lead of 53-47, up from 51-49 in April. The forced preference primary vote results are Labor 39.6%, Liberal 36.6%, Greens 7.8% and others 16.0%. On previous election preferences, Labor’s lead would be still greater, at 54.7-45.3. Karen McNamara won the Central Coast seat for the Liberals in 2013 by 0.7%, but the redistribution has left it with a notional Labor margin of 0.2%. Polls earlier in the campaign showed very little in it: a ReachTEL poll for the Fairfax papers on June 9 had it at 51-49 to Labor, and a Galaxy poll for the News Corp tabloids on May 11 had it at 50-50.

Lindsay (Liberal 3.0%): Four earlier polls, including two from ReachTEL and two from Newspoll/Galaxy, showed the Liberals leading in the outer western Sydney seat, but the NSWTF’s poll of 610 respondents has Labor’s Emma Husar with a 54-46 lead over Liberal member Fiona Scott. Primary votes are Labor 39.9%, Liberal 34.5%, Greens 3.9% and others 21.6%. On previous election preferences, Husar’s lead is 54.6-45.6. Scott led 53-47 in a Newspoll on June 14, 51-48 in a ReachTEL poll for the Australian Education Union on June 13, 54-46 in a ReachTEL for Fairfax on June 9, and 54-46 in a Galaxy poll for the News Corp tabloids on 54-46.

Macquarie (Liberal 3.3%): The poll of 636 respondents has Labor candidate Susan Templeman with a 54-46 lead over Liberal member Louise Markus in the western Sydney hinterland electorate, from primary votes of Labor 38.9%, Liberal 38.9% and Greens 12.1%. On previous election preferences, Templeman’s two-party lead is 55.0-45.0.

Eden-Monaro (Liberal 2.9%): In a poll of 636 respondents in the famous bellwether electorate in the state’s south-eastern corner, Labor’s Mike Kelly has a commanding lead of 55-45, or 54.0-46.0 on previous election preferences. Primary votes are Liberal 40.4%, Labor 37.5%, Greens 14.8% and others 7.3%. A similar poll for the Australian Education Union on June 18 produced much the same result.

Gilmore (Liberal 3.8%): A poll of 632 respondents in the southern New South Wales seat, which was last held by Labor in 1996, has Labor’s Fiona Phillips leading Liberal member Ann Sudmalis by 53-47. Primary votes are Liberal 39.3%, Labor 37.0% and Greens 12.5%. The result on previous election preferences is much the same (52.7-47.3). A poll by Galaxy for the News Corp tabloids on May 11 had the Liberals ahead 51-49.

Page (Nationals 3.1%): A poll of 647 respondents has Labor’s Janelle Saffin leading 54-46 in the seat she lost to Nationals member Kevin Hogan in 2013, from primary votes of Nationals 39.1%, Labor 36.6% and Greens 15.4%. On previous election preferences, the two-party difference is 53.3-46.7. A ReachTEL poll for the Australian Education Union on June 13 had Saffin leading 52-48.

Elsewhere:

• James Massola of Fairfax offers an account of the state of play based on discussion with “more than a dozen Labor strategists, officials, MPs and campaign workers across every state of Australia on Wednesday – as well as Liberal and National party strategists”. This suggests Labor has 66 seats pencilled in, with a good deal many more not being written off, and a hung parliament being well within the range of possibilities.

• The West Australian reports Labor polling “has picked up substantial swings against the coalition in Burt, Cowan, Swan and in the safe Liberal seat of Pearce, held by Cabinet minister Christian Porter”. Laurie Oakes related on the weekend that Labor had detected a 9% swing in Pearce, where Porter’s margin is 9.3%, and I gave the pot a further stir in a paywalled WA situation report in Crikey yesterday.

• According to Sharri Markson of The Australian, polling for the Nationals confirms the findings of a ReachTEL poll for GetUp! on June 13 in showing independent Rob Oakeshott on over 20% of the primary vote, leaving him “neck and neck” with Nationals member Luke Hartsuyker after preferences.

Six weeks into the campaign, the BludgerTrack poll aggregate is where it was at the start of the campaign, and has been at every point since – with nothing at all to separate the Coalition and Labor on two-party preferred. The only changes since a week ago have occurred at state level, where the Coalition is down two on the seat projection in New South Wales, but up one each in Queensland and Western Australia. The latest addition to the aggregate is your regular weekly Essential Research, which is unchanged on two-party preferred at 51-49 in favour of Labor. On the primary vote, the Coalition is down one to 40%, while Labor, the Greens are the Nick Xenophon Team are steady on 37%, 10% and 4% respectively. Speaking of the Nick Xenophon Team, it should be noted that the BludgerTrack model ignores its existence so far as the seat projection is concerned, so the following should probably be interpreted as pointing to a hung parliament:

There’s a bunch of electorate-level internal polling doing the rounds at the moment, something that always needs to be viewed with regard to the fact that those who commissioned might only be publicising the results that they like. Nonetheless, the display at the bottom of this post, which is updated with all the latest results, shows up no distinction in the average swing recorded across media and private polls over the course of the campaign period. As for published polling, Essential Research should, as usual, be with us later today. Roy Morgan has decided to dispense with its national polling and instead focus on electorate-level polling for the remainder of the campaign, the latest example of which isolates the ten strongest seats for the Greens. These results are based on samples of around 300 aggregated from all the outfit’s regular polling going back to January. That means a good deal of the survey period was from a time when the Coalition still had a substantial lead, and the “others” vote was lower than it has since become. Morgan has presumably, and probably correctly, concluded that it will generate more headlines this way than if it were merely one national poll among many.

• The Age reports that a poll conducted for the Greens suggests the party to be well in the hunt in the inner Melbourne seat of Batman, despite the blow dealt them when the Liberals announced its how-to-vote cards would preference Labor ahead of them. The poll has Greens candidate Alex Bhathal leading Labor member David Feeney by 41% to 28% on the primary vote, which pans out to 55-45 on respondent-allocated preferences, and would produce much the same result on 2013 election preferences. The automated phone poll was conducted by Lonergan Research from a large sample of 1600 respondents. However, The Age report also relates that “internal and larger-scale polling for the ALP” actually shows Labor leading on the primary vote. The report also says Labor’s poll shows the party to be “much more popular with voters under 24 than the Greens”, whose “strongest age bracket is 35-50 year olds” – a finding that frankly isn’t credible.

• The Advertiser reports that “high-profile Labor frontbencher Kate Ellis is facing a shock defeat to the Liberals in her inner-city seat of Adelaide”, although it’s based on a rather thin sample of 364. The poll was conducted for the ALP by ReachTEL, which I’ve never seen associated with a sample of this size before (UPDATE: And sure enough, ReachTEL denies it was their poll). According to the report, the poll credits Liberal candidate David Colovic with a 51-49 lead over Labor member Kate Ellis. The report speculates that Labor provided the paper with the polling “to rally support for Ms Ellis in the face of a statewide surge by the Nick Xenophon Team”.

• The Australian Education Union is circulating three ReachTEL polls conducted in marginal seats in New South Wales, one of which shows Labor with a commanding lead of 55-45 in the legendary bellwether electorate of Eden-Monaro. Primary votes are 41.2% for Liberal incumbent Peter Hendy, 38.6% for Labor challenger Mike Kelly, and 11.0% for the Greens. A fairly extraordinary flow of respondent-allocated preferences pushes Labor’s two-party total well past where it would be based on 2013 election preferences, in this case 52.6%. Sample: 719.

• In Lindsay, Liberal member Fiona Scott has a narrow lead of 51-49 over Labor candidate Emma Husar, the primary votes being Liberal 42.9%, Labor 36.6%, Christian Democratic Party 6.4% and Greens 5.3%. Based on previous election preferences, Scott’s lead is 51.6-48.4. Sample: 656.

• In Page, Labor challenger Janelle Saffin leads Nationals member Kevin Hogan 52-48 on both respondent-allocated and previous election preferences, from primary votes of Nationals 42.1%, Labor 38.4% and Greens 12.2%. Sample 788.

• The Daily Telegraph reports that a poll of the South Australian regional seat of Barker, conducted by ReachTEL for the CFMEU from a sample of 869, has Nick Xenophon Team candidate James Stacey leading Liberal incumbent Tony Pasin by 52-48.

• A ReachTEL poll for the eastern Melbourne electorate of Menzies, conducted for independent candidate Stephen Mayne, credits Liberal member Kevin Andrews with a two-party preferred vote of 61-39, which would be 63-37 on previous election preference flows. Andrews’ share of the two-party vote in 2013 was 64.4%. The poll was conducted June 13 from a sample of 719.