Saturday, 18 May 2019

House of Representatives Predictions


Voting has already opened in all states and territories, and I have yet to post a prediction. This isn’t the first time, but it is later than I would like and means the predictions this year are going to be just the bare basics and not consider the Senate races.

The Barest of Basics

As of this election, Australia has 151 seats. This is one more than previously with the addition of the Seat of Bean and means that there has been some significant reshuffling of boundaries. For this reason, the pendulum—a list of seats and how close they were as a two-party contest last election/by-election—is based on mathematically adjusted data. 140 of these were simple Coalition vs Labor contests as shown below. The ‘Current Coalition Vote’ represents the share of the 2-party vote after mathematically adjusting from past elections.


We can easily make a general prediction by applying a uniform swing to these seats. In the 2016 election, the Coalition won 50.4% of the two-party vote while the latest Newspoll has them down to 48.5%. This is a drop of 1.9 percentage points, which doesn’t seem like much but would see the Liberals lose 10 seats (Banks, Corangamite, Capricornia, Dickson, Dunkley, Flynn, Forde, Gilmore, Petrie and Robertson), including the two Victorian seats already lost in the mathematical adjustment for border changes (Corangamite & Dunkley).


This assumes a uniform, national swing and that there isn’t a chance for a minor party to win in any of these seats. It also ignores that fact that polling since 2016 isn’t a perfect baseline where more recent by-elections have been held, and we’ve had an unusually high number of those. On these numbers alone, though, Labor would win 77 seats, more than the 75 needed to form a government. This is without even considering the possibility of winning some of the remaining 11 seats that are not simple Coalition vs Labor fights (e.g. Cooper, Wills and Grayndler which are held by Labor but their primary opponent is the Greens party.) Being 14 seats behind at this point, even a minority Liberal government looks highly implausible.

Initial Conclusion: ALP Majority Government.

Slightly Better Dressed Basics

My gut feeling is that Labor’s climate change policy and a protest against Liberal support for Adani will see the ALP do even better in inner Melbourne and Sydney, but also that these seats will have a stronger-than-usual Greens vote. Conversely, Labor may struggle a little more in Queensland because of the jobs promised by Adani, and half of the ten seats mentioned above (i.e. Capricornia, Dickson, Flynn, Forde and Petrie) come from that state. Petrie and Dickson, in particular, are held by the Liberal National Party by 1.6 and 1.7 percentage points respectively so any erosion of the predicted swing of 1.9 could see the Coalition hold on here.

For simplicity, I’ll use the Electoral Commission’s maps of Inner Sydney and Melbourne to define the seats where Labor may do better than average: Aston, Banks, Barton, Bennelong, Blaxland, Bradfield, Bruce, Calwell, Chifley, Chisholm, Cooper, Deakin, Dunkley, Fowler, Fraser, Gellibrand, Goldstein, Grayndler, Greenway, Higgins, Holt, Hotham, Isaacs, Jagajaga, Kingsford Smith, Kooyong, Lindsay, Macnamara, Maribyrnong, McMahon, Melbourne, Menzies, Mitchell, North Sydney, Parramatta, Reid, Scullin, Sydney, Warringah, Watson, Wentworth and Wills.

By standard definitions, a seat is ‘safe’ if it has a margin of over 6% (i.e. is 56% or more for the Coalition in the above tables, or 44% or less for Labor). Removing currently safe seats from this list and those already held by Labor leaves a more manageable list of Banks, Chisholm, Dunkley and Reid for consideration. Banks and Dunkley would become ALP seats even if Labor only achieves a swing of 1.9, so let’s call them Labor wins already. On the other hand, I’ll keep in Warringah, as there has been an active campaign against Tony Abbott in his seat here.

This gives us a nice short list of seats that require a little extra thought:

  • Capricornia, Dickson, Flynn, Forde and Petrie, which may resist Labor’s swing out of uncertainty how the ALP will react to Adani;
  • Chisholm and Reid, where the ALP should do well, and the Greens may be an outside chance;
  • Warringah, where Tony Abbott is unpopular;
  • Clark, Indi, Kennedy, Mayo, Melbourne and Wentworth, which aren’t in the tables above as independents or minor parties currently hold them;
  • Cooper, Grayndler, Higgins and Wills, also not in the tables above and where the Greens are the main threat to the incumbent; and
  • Cowper, which in 2016 was a race between the Nationals and high-profile independent Rob Oakeshott and is not in the tables above.

The remaining seats fall 60 to the Coalition and 72 Labor, which already looks pretty conclusive.

Basics in Formal Attire

Capricornia

With Forde, Capricornia is tied for the most marginal Coalition seat in Australia. It also has a long Labor history, with the ALP only losing two elections between 1972 and 2013. Mining jobs aside, if any seat were to fall to Labor it’d be one of these two. Capricornia is still more agricultural than mining, and I doubt Adani’s promise of jobs will be enough to save either of these seats.

Prediction: ALP (possible)

Chisholm

Nestled in Melbourne’s inner east, Chisholm’s vote for the Liberals in 2016 can be seen as something of an aberration, with Labor holding the seat before this as far back as 1998. This seems to have been a pro-Turnbull move, and the incumbent Liberal is not contesting the seat because of Turnbull’s knifing. 2.9% isn’t a big margin, only slightly larger than the polled national swing, inflated by support for a fired PM and contesting-elsewhere-as-an-independent MP. It seems like a correction is inevitable.

Prediction: ALP (likely)

Clark


Independent Andrew Wilkie holds Clark by a massive margin of 17.8. He will easily retain this seat.

Prediction: IND (v likely)

Cooper

Labor holds Cooper in a close race with the Greens. First preference voting has slightly worsened for the Greens nationally, but Melbourne is a hotspot for opposition to Adani, so I think they’ll do better than last time. On the other hand, I think the Greens have a good shot here and had the highest primary vote. Liberals are not a real contender in Cooper, and whether Labor or Liberal wins it will still be a vote for an ALP government.

Prediction: GRN (likely)

Cowper

Cowper has been a Nationals seat since before the National Party was called the National Party. However, Rob Oakeshott is a well-known independent and has served as a member of parliament for a previous seat. The current Nationals’ candidate is retiring, leaving a relative unknown to contest the seat, and the Nationals in NSW have been rocked by accusations of corruption around Water in the Murray Darling and the activities of Barnaby Joyce. Add to that the opportunity for conservative voters the option to vote against the Nationals without supporting Labor, and things are looking good for the independent.

Past voting shows that Rob Oakeshott won’t be able to win on primary votes alone. Last election he only received 26.3% compared to the Nationals’ 46.0%. Even Labor and Greens preferences weren’t enough to get above the Nats’ primary vote, but it got close.
Cowper will see Oakeshott beat Labor and Greens, and scoop up most of their preferences (probably even more strongly than in 2016. The question is whether this will be enough. That’s a really tough call, but I think it probably won’t.

Prediction: NAT (dubious) 

Dickson

Sitting on a 1.7% margin, the seat would fall to a uniform, national swing of 1.9 to the ALP. It is receiving further consideration for possible support for Adani, but Dickson is held by Peter Dutton who is hugely unpopular here for his role in removing Turnbull. So…

Prediction: ALP (possible)

Flynn

Flynn has a margin of 1.0%, but in central Queensland there is the prospect of jobs from Adani and uncertainty around how Labor will affect this. If the Adani issue were to save any Coalition seat in Queensland, it’d be Petrie. If it were to save two, the second would be Flynn. On that basis, I’m very unsure.

Prediction: ALP (dubious)

Forde

With Capricornia, Forde is tied for the most marginal Coalition seat in Australia. Mining jobs aside, if any seat were to fall to Labor it’d be one of these two. Forde is more Coalition-leaning historically than Capricornia, but I doubt Adani’s promise of jobs will be enough to save either of these seats.

Prediction: ALP (possible)

Grayndler

The Liberals have no chance in Grayndler. Labor has won the seat every election since 1972, and in 2016 had a primary vote twice that of the Liberals. Greens and Liberals polled quite similarly in 2016, so preferences are all important here. If Liberals stay ahead of the Greens when all minor parties are ruled out, Greens voters will flow to Labor to create one of the safest seats by margin in the country (somewhere around 70:30 ALP vs Lib). If the minor parties help the Greens to out-poll the Coalition, it’s more complicated. Anyone-but-Labor voters will boost the Greens, but most Liberals favour the centre-left ALP over the left-left Greens. This is what happened in 2016, causing the ALP v Greens race that makes this seat worth a second look. In that race, Labor was a clear winner with a margin over the Greens of more than 15%.

Prediction: ALP (v likely)

Higgins

The Liberals hold this seat by 7.4%, making it safe. In fact, they’ve held the seat since ’72. The Greens did well here, beating Labor, but the Liberals won on a primary vote of 51.6%, and the magnitude of swings required to see anyone else win is radical.

Prediction: LIB (v likely)

Indi

Indi has been coalition-held since 1972 if you ignore the incumbent independent, Cathy McGowan. And we will ignore her for now, as she is retiring. Second in 2016 were the Liberals, and then the Nationals with Labor only polling 10.1%. It is almost unheard of for independents to succeed independents. McGowan’s personal brand is just that, personal, and I’d expect this to resolve easily back to the Coalition.

Prediction: LIB (v likely)

Kennedy

Bob Katter has held Kennedy since 1993, and with a margin of 11%, he isn’t going to stop now.

Prediction: KAP (v likely)

Mayo

Rebekah Sharkie (Centre Alliance) beat Georgina Downer (Liberals) with a narrow 2.9% margin but has a reasonably high satisfaction in Mayo. Enthusiasm for a return of the Downer line and anger about Sharkie’s dual-citizenship situation in the 2018 by-election may have waned slightly, particularly with the general swing against the Libs. After accusations that Georgina had been “parachuted in” last by-election, the Libs couldn’t seriously propose any local candidate as the best choice in town without them being seen as even worse than Downer. Parachuting in another candidate would obviously backfire too. That leaves the Libs stuck with Downer, and Downer stuck with Mayo this time around. 2022 may be different, but this once safe Liberal seat has probably been conceded by the party already.

Prediction: CA (v likely)

Melbourne

Melbourne is the Greens’ only seat. They hold it against the Liberals who narrowly out-polled the ALP 24.8% to 24.0%. The predicted swing to Labor should get the Labor party to second place, and Liberal preferences will decide the race rather than Labor’s. This could massively undermine the Greens here. On the other hand, being the most prominent Greens candidate in the House of Reps races has its perks, and Adani protests are big here. Greens retaining the seat is certainly not out of the question.

Prediction: GRN (dubious)

Petrie

With the LNP sitting on a 1.6% margin, a uniform, national swing of 1.9 to the ALP is a threat. The generally Liberal trend in past votes is a consideration in the Liberal Nationals’ favour. If the Adani issue was to save any Coalition seat in Queensland, it’d be Petrie. On the assumption that at least one seat will be influenced by the coal mine…

Prediction: LNP (possible)

Reid

Reid’s Liberal is sitting on a 4.7% margin, which is almost safe, and a history of Labor voting from 1972 until 2013, which is very much not safe at all. Adani, Turnbull and other issues are likely to weigh heavily here.

Prediction: ALP (possible).

Warringah

Warringah is only on this list because of anger against Tony Abbott and a concerted campaign to cost him his seat even among traditional Liberals. With 11.1% as a margin, this shouldn’t be in play, but it is. However, there are two independents, Fraser Anning’s Conservative Nationals, United Australia and the Christian Democratic Party to scoop up protest votes and dissipate anger before preferences flow back to the Liberals. I’m expecting Abbott will hold this seat, but it is uncertain. Though unlikely in my opinion, this is also probably United Australia’s best shot at a lower house seat.

Prediction: LIB (dubious)

Wentworth

Wentworth was won in a by-election by independent Dr Kerryn Phelps after Malcolm Turnbull resigned, but after boundary changes is suggested to be 51% Liberal. This is probably doubtful, as the new inclusions in the seat didn’t have the option of voting for Phelps previously, and thus their attitudes aren’t properly reflected in their past voting data. Additionally, 1% wouldn’t be enough to win in a contest against Labor.

Anger about the ousting of Turnbull has probably lessened, but this looks like a comfortable seat for Phelps.

Prediction: IND (likely)

Wills

Excepting the 1992 by-election and 1993 general election appointing an Independent, Wills has been won by Labor in every contest since 1972. Last year was a close race against the Greens which complicatedly picked up a lot of Liberals’ votes over Labor. The ALP’s margin here is only 4.9 and Adani will likely tighten this in Melbourne’s north-west. It seems the long Labor history has made Liberals in this seat willing to side with Greens, and that makes this a tough seat to get a feel for. I’m going to play it safe and stick with what I know.

Prediction: ALP (dubious)

Conclusion

Factoring these predictions into our broader table, we get the final prediction for the House of Reps:


Prediction: ALP Majority Government

ALP: 80
Coalition: 65       (LIB: 38; LNP: 17; NAT: 10)
Cross-bench: 6   (GRN: 2; CA: 1; KAP: 1; IND: 2)

No comments:

Post a Comment