Category: Federal Redistributions

The proposed federal redistribution for New South Wales, requiring the abolition of one of its 47 seats, is being published today. The boundaries will presumably be up on the Australian Electoral Commission site shortly, but for now there a gazetted notice informs us that the proposal abolishes North Sydney, held by teal independent Kylea Tink. The fact that no other electorate names are identified as having been abolished, and the revelation that More to follow.

UPDATE: Full boundaries now available on the AEC site, and here are my two-candidate and two-party preferred (the latter boiling it down to Labor and Coalition) estimates for the new margins. Teals are treated as a single entity so, for example, where Warringah gains territory from North Sydney, Kylea Tink’s votes there are transferred to Zali Steggall. In areas where there was no teal candidate last time though, it’s not possible to estimate how they would have gone.

UPDATE 2: I just unearthed an error in the code that was causing errors in a few places, notably Hume and Eden-Monaro. The numbers below should add up now.

UPDATE 3: My analysis of the new boundaries has just been published in Crikey.

Proposed new federal boundaries have been published today for both Western Australia and Victoria.

Victoria

The Victorian proposal is here. The proposed seat for abolition is Higgins, reflecting under-enrolment in the inner eastern areas of Melbourne. Below are my estimates of the new vote shares (you’ll have to click on the image for it to be legible), which can be compared with Ben Raue’s similar exercise. Note that there are three seats where the two-candidate preferred has to be split three ways: Melbourne is to take a chunk of Higgins, for which there are no ALP vs GRN numbers; Goldstein gets parts of Hotham and Isaacs, where there is no TEAL vs LIB; and Kooyong gets a big piece of Higgins, ditto.

The best news for Labor is that Menzies now has a notional Labor margin of 0.7% (0.3% by Ben Raue’s reckoning) compared with an actual Liberal margin of 0.7%. Beyond having lost Higgins, the news is bad for them in Chisholm, where their margin is down from 6.4% to 3.3% by Ben Raue’s reckoning and 2.8% by mine, and Wills, where the Greens have been handed the gift of Carlton North and Fitzroy North from Melbourne, cutting the Labor margin over them from 8.6% to 4.2%. However, their 0.7% improvement in Dunkley might prove handy one day.

At first glance, I would imagine that Monique Ryan would be pleased not merely to have had her electorate maintained, but to have had it supplemented with the inner urban end of Higgins and to have lost territory in the east to Menzies and Chisholm. Goldstein teal MP Zoe Daniel’s gains from Hotham and Isaacs are perhaps less helpful, though that is harder to read.

Western Australia

The proposed sixteenth seat for the state is called Bullwinkel and encompasses eastern suburbs of Perth and the Avon Valley further afield, taking much of the territory of Hasluck along with parts of Swan, Burt, Canning, Durack and O’Connor. The new seat has a notional Labor margin of 2.9%, but in the absence of a defending sitting member and given the unusual nature of the 2022 result, the Liberals would be favoured to win. However, its creation has given Labor a helpful 4.7% boost in Hasluck, which loses its most conservative territory to the new seat. Andrew Hastie is a loser out of the redistribution in Canning, but is presumably not in too much danger. Labor has been favoured slightly in Tangney and Swan, Liberal in Cowan. The changes are unlikely to make much difference to teal independent Kate Chaney in Curtin.

In a remarkable display of synchronicity, yesterday saw the announcement of finalised new federal electoral boundaries for Victoria, which will be followed today by the publication of draft new boundaries for Victorian state boundaries, the result of an entirely separate process. While the former release is limited to written descriptions of the changes that have been made to the draft that was published in March, with maps and geospatial data not to be published until July 28, I have taken the effort to conduct a full analysis of the final boundaries, which you can read about below. I hope to be able to provide an analysis of the state boundaries later (probably much later) today.

The federal redistribution gives effect to an increase in Victoria’s representation from 38 House of Representatives seats to 39, at the expense of Western Australia, which is down from 16 to 15 and has accordingly undergone its own redistribution which was finalised a fortnight ago. The final boundaries in Victoria make changes to the draft boundaries affecting 20 electorates, by far the most significant of which is the abandonment of a major adjustment that was proposed for the boundary between Macnamara and Higgins in Melbourne’s inner south, an area of particular interest to the Greens.

N.B.: I am a little confused by the assertion of the AEC release that a remnant of the change remains in that most of Windsor has been transferred to Higgins, as it seems to me that both draft and final boundaries have this suburb remaining entirely in Macnamara. This analysis proceeds on the assumption that this statement is in error, but I may well be missing something. Ben Raue’s maps of the draft boundaries at The Tally Room are instructive here. (UPDATE: This has finally penetrated my skull — Windsor will indeed be transferred to Higgins, with effects that do little to disturb my overall analysis. The spreadsheet linked to at the bottom of the post has been revised to reflect this).

The practical upshot of the reversal is that the Greens are now less likely to take Macnamara from Labor, but the chances of them rather than Labor taking Higgins from the Liberals has increased, without much affecting the Liberals’ overall chances of losing the seat. The draft proposed a straight north-south boundary along Williams Road, resulting in Macnamara gaining territory around South Yarra at the northern end, while an area around Caulfield in the south was to go the other way. This would have boosted the Greens in Macnamara by around 2% from their 24.2% at the 2019 election, drawing in roughly equal measure from the Liberal and Labor vote. Just a few days ago, this prospect had the Greens talking up their chances of taking the seat from Josh Burns, who retained it for Labor when Michael Danby retired in 2019.

Macnamara, which was known prior to the 2019 election as Melbourne Ports, has been in Labor hands since 1906, but has lately evolved into a tight three-way contest involving the Liberals, who came within 2.2% of winning the seat in 2016, and the Greens, who would almost certainly win if they were able to reduce Labor to third place. This they fell 5.8% short of doing so in 2019, when they were excluded at the second-last count with 27.33% of the vote to Labor’s 33.2%. Greens preferences then flowed heavily enough to Labor to secure a comfortable 6.2% margin for Josh Burns, despite the Liberals outpolling him by 37.4% to 31.8% on the primary vote. The changes proposed by the draft boundary would have brought the Greens’ shortfall to around 3%, in a seat where their primary vote had steadily escalated from 15.0% to 24.2% since 2007.

However, the Greens’ gain in Macnamara would have been balanced by a weakening in Higgins, which on my analysis would have cut their primary vote by 1.6% while boosting Liberal and Labor by 0.7% each. Higgins has also developed into a three-cornered contest in recent years, a particularly notable fact in a seat that the Liberals have held since its creation in 1949, with members including John Gorton and Peter Costello. The Greens came closer to taking second place in Higgins in 2019 than they did in Macnamara, being excluded from the count with 24.3% to Labor’s 26.2%. However, the Liberals only fell just short of a majority at this point with 49.5%, and would have picked up enough stray preferences after the final exclusion to have retained the seat in any case.

Preference flows in Melbourne’s inner urban seats in 2019 suggest the Liberals would be slightly more likely to lose Higgins if Labor rather than the Greens made the final count, since flows from the Greens to Labor were slightly stronger than vice-versa — particularly in Macnamara, where Greens preferences split 87.5-12.5. However, the distinction seems to blur in areas where the Liberal vote is stronger, with Greens preferences in Higgins splitting 83.4-16.6, which was equal to how Labor preferences split between the Greens and the Liberals after they finished third in Kooyong.

The only other change to the finalised boundaries that might potentially have a bearing on the election result involves Chisholm, Higgins’ eastern neighbour, which Gladys Liu retained for the Liberals in 2019 by 0.6% after a 2.3% swing to Labor. The seat will be reoriented southwards with the redistribution, which by my reckoning added 0.2% to the Liberal margin on the draft boundaries — although Antony Green’s calculation was that it had in fact gone 0.4% the other way. My assessment is that the minor adjustments made in the final boundaries boost the Liberals by a further 0.3%.

You can find my party vote share estimates for the new boundaries, together with detailed accounting of how they were arrived at, here for Victoria and here for the Western Australian redistribution.

Federal redistributions for Victoria, South Australia and the Australian Capital Territory have been finalised over the past fortnight, a development that removes procedural obstacles for the staging of a normal election for the House of Representatives and half the Senate. At the bottom of this post is a new electoral pendulum based on post-redistribution margins. This illustrates that the goverment has, notionally speaking, lost its majority, being reduced from 76 seats to 74 in a chamber that increases in size from 150 to 151. The Liberal-held Victorian seats of Corangamite and Dunkley move into the Labor column (just barely in the former case – others who calculate the margins might very easily fall the other way), while Labor also gains new seats in Victoria and the Australian Capital Territory, but loses a seat through the abolition of Port Adelaide in South Australia. That gives Labor 72 seats, assuming the cross-bench remains unchanged with its existing five members. All margins shown in the pendulum are Labor-versus-Coalition, except in the five seats held by minor parties and independents.

The main changes made from the draft to the finalised boundaries are cosmetic, in that Corangamite will retain that name and not be renamed Cox, and Batman will be renamed Cooper. Beyond that, minor changes were made to eleven seats in Victoria. The most significant change was to Deakin, costing it around a third of the 19,000 voters it was originally to gain from Casey in the east, and compensating it at the other end with more than 3000 extra voters in Vermont South from Chisholm. The electoral impacts are very slightly advantageous to the Liberals, boosting their margin in Dunkley relative to the draft proposal by 0.4% to 6.6%. There are no changes to the draft as far as I can see in South Australia, and only a minor and inconsequential one in the Australian Capital Territory.

The links below provide my full accounting of the new margins, both for the three redistributions just finalised, and the other three concluded since the last election.

Federal Redistribution of Victoria 2018
Federal Redistribution of Australian Capital Territory 2018
Federal Redistribution of South Australia 2018

Federal Redistribution of Queensland 2018
Federal Redistribution of Tasmania 2017
Federal Redistribution of Northern Territory 2017

The two-party results featured above are strictly Labor-versus-Coalition, which leaves some explaining required where this doesn’t apply. None of the three Melbourne inner-city seats where the Greens threaten Labor has been significantly changed. There is no difficulty counting a new Labor-versus-Greens result in Wills and the seat now known as Cooper, since neither gains territory from a seat in which no Labor-versus-Greens count was conducted. David Feeney held the Greens off in Batman by 1.0% in 2016, but I now make it at 0.5%, while Ben Raue at The Tally Room has it at 0.7%. However, given Ged Kearney’s succession to the seat in the March by-election, comparisons based on the 2016 election are rather academic. Labor’s margin over the Greens in Wills is unchanged at 4.9%. Cathy McGowan’s seat undergoes only minor changes, although territory gained from Murray (now Nicholls) now provides about 4% of its voters, which obviously cannot be used to measure support there for McGowan. This is not enough to significantly alter the position of McGowan, who held off Sophie Mirrabella in 2016 by 4.8%.

In South Australia, the Nick Xenophon Team reached the final count in four seats, winning Mayo from the Liberals by a margin of 5.0%, and finishing 2.0% and 4.7% short of Liberal incumbents in Grey and Barker, and 14.9% short of Labor in Port Adelaide. In no case can new margins be determined: Port Adelaide is abolished; Grey and Barker have both gained territory from Wakefield, respectively accounting for around 15% and 8.5% of their voters, where no Liberal-versus-NXT count was conducted; and Mayo gains territory from Kingston and Boothby, collectively accounting for over 17% of the voters, neither of which had useable counts.

Pendulum over the fold below. Seats whose notional party has changed are indicated with an asterisk. The “redist.” column records the effect on the margin of redistributions, where they have been conducted (i.e. everywhere but New South Wales and Western Australia). Margins shown for the five seats held by minor parties and independents are their winning margins at the 2016 election, with no change made for the redistribution.

Continue reading “Finalised redistributions and federal election pendulum”

The Australian Electoral Commission has published draft boundaries for the South Australian redistribution, which brings the state’s representation down from 11 seats to 10. The seat mooted for abolition is Port Adelaide; Wakefield is to be renamed Spence. At the bottom is a table featuring my estimates of party vote shares and two-party margins (Labor versus Liberal only).

Adelaide. Drifts westwards into the void created by the abolition of Port Adelaide, turning a tight Labor seat into a reasonably safe one.

Barker. Gains Barossa Valley territory around Kapunda.

Boothby. Drawn northwards into Glenelg through knock-on effects from Port Adelaide abolition, without much change to the margin.

Grey. Expends to the northern edge of Adelaide, gaining Clare Valley, with next to no impact on the margin.

Hindmarsh. Takes the bulk of Port Adelaide, turning the seat from marginal to safe Labor.

Kingston. Loses coast at southern end around Aldinga Beach, gains suburbia at northern end around Aberfoyle Park. Slightly advantageous to Liberal, but not enough to make them competitive on recent form.

Makin. Expands west to take over some Port Adelaide territory, notably Parafield Gardens.

Mayo. Not abolished, as some expected; gains the Aldinga Beach coastal area lost by Kingston.

Spence (Wakefield). Greatly strengthened for Labor through loss of Clare Valley and Barossa Valley to Grey and Barker respectively, and gain of suburbs around Paralowie from Port Adelaide.

Sturt. Gains Norwood at western end from Adelaide, with little impact on margin.