Assuming this parliament runs its full term, there are likely to be redistributions in every state except South Australia before the next election, as well as the Northern Territory. A Western Australian redistribution is currently in its early stages, resulting from the rule requiring that states and territories be redistributed at least every seven years. Tasmania and the Northern Territory will follow for the same reason later this year; Victoria is also due in January 2010 (UPDATE: Or possibly not - see David Walsh in comments). It is also very likely that population changes will result in New South Wales losing another seat to Queensland when the determination is made in February 2009. The following table shows the states’ entitlements at the last determination in November 2005, entitlements based on ABS population figures from June 2007, and a projection to February 2009 based on growth rates in the 2006/07 financial year. Note that the constitution prevents Tasmania from falling below five seats, and that a dubious law passed in 2004 allows the Northern Territory to retain its second seat even if its entitlement falls a little below 1.5.

In the meantime we have a redistribution in train for Western Australia’s existing 15 seats, which despite the state’s rapid growth will not need to increase before the next election. Population volatility has led to substantial variations in enrolment across the electorates, with growth trends confounding the projections used to conduct the last redistribution in 2000. The following table shows actual enrolment at that time; the projections then arrived at for May 2004, three-and-a-half years hence; actual enrolment figures from the October 2004 election; and enrolment as of last month.

It is particularly notable that Moore, heretofore the quintessential growth corridor electorate, has fallen well short of the AEC’s projections at the time of the 2000 redistribution. Enrolment in the electorate took seven years to grow 6.9 per cent, against a projected 17.2 per cent in three-and-a-half years. This is a similar rate of growth to other Perth suburban seats, which came in between 1 per cent and 10 per cent. The real action has been in semi-rural Canning (32.5 per cent), Pearce (29.2 per cent) and Brand (24.8 per cent), along with outer metropolitan Cowan (22.0 per cent). Growth in the state’s south-west has boosted enrolment in safe Liberal Forrest by 19.6 per cent, but further afield O’Connor and Kalgoorlie have remained stagnant.

It will thus be necessary for the redistribution to cut upwards of 10,000 voters from Canning, Pearce, Brand, Cowan and Forrest, paring back existing over-enrolment and accounting for projected growth over the next three-and-a-half years. Significant expansion will be required not only for ever-declining Kalgoorlie and O’Connor, but also for Moore to correct for its over-estimated growth prospects last time. In the metropolitan area, Curtin, Swan and Tangney will need to take in new areas, but little adjustment will be necessary for Perth, Stirling and Fremantle (which is not to say that these electorates will not be redrawn due to knock-on effects).

While it never pays to second-guess redistributions, it’s tempting to draw a scenario in which Kalgoorlie absorbs all or part of Geraldton from O’Connor, which can easily be compensated by taking some of the territory that Pearce and Forrest need to lose. The need for cuts to adjoining Forrest, Canning and Brand will tempt the commissioners to make most of this required transfer from Forrest, resulting in knock-on transfers to the other two. The required growth in Moore can be accommodated either at the expense of Pearce to the north or Cowan to the east, both of which will need to be cut.

A few items worth noting from the Queensland state scene:

• The state’s Electoral Commission has published preliminary submissions for the redistribution currently under way. The proposed redistribution report is still a few months away.

• Brisbane’s Sunday Mail has published an interesting Galaxy poll of 800 respondents showing Anna Bligh with a 68 per cent approval rating, compared with just 12 per cent for Nationals leader Jeff Seeney. The poll puts Labor’s two-party preferred lead at 55-45. It also found Mal Brough favoured by 32 per cent to lead the state Liberals against 8 per cent each for the recently departed Bruce Flegg and his replacement Mark McArdle, and 7 per cent for thwarted contender Tim Nicholls. McArdle doesn’t seem too upset at the prospect of being replaced by Brough, though it’s far from clear whether Brough is really interested in state politics.

• A recent Queensland Newspoll got overlooked in my earlier post on state results, so here’s a chart showing the two-party preferred score going back to 2004. I have gone for a time-series X axis on this occasion, which is messier but more accurate.

To celebrate today’s primaries in Michigan, I hereby present presidential election open thread number two.

For the most part, the November 24 federal election saw Labor pick up swings of similar magnitude in Victoria across city and country seats, notwithstanding the unusually small swings in the inner-city. However, one intriguing variation with respect to country Victoria was a relatively weak Labor performance east of Melbourne, particularly in the West Gippsland electorate of McMillan.

McMillan MP Russell Broadbent enjoyed one of the best evenings of his chequered electoral career by suffering a swing of just 0.2 per cent, a result which echoed the surprise Coalition wins in the local seats of Narracan and Morwell at the November 2006 state election, following respective swings of 9.5 per cent and 7.0 per cent (the defeated Labor member for Narracan, Ian Maxfield, is the husband of Labor’s unsuccessful McMillan candidate, Christine Maxfield). These electorates correspond with the rural areas of McMillan which gave Broadbent his best results: the small booths here gave early indications on election night of a swing in Broadbent’s favour, providing the Liberals an unwarranted early morale boost. However, the three large Pakenham booths which came in later followed the outer suburban script, swinging to Labor by around 7 per cent. There were also relatively strong performances for the Coalition in the neighbouring electorates of Flinders and Gippsland. Flinders followed the pattern of McMillan in that the biggest swings were recorded in the urban fringe centre of Somerville. Peter McGauran defied occasional talk about the long-term security of his hold on Gippsland by holding on to all but 0.4 per cent of his primary vote, suffering an evenly distributed 1.8 per cent two-party swing.

Labor enjoyed stronger performances to the west and north of Melbourne, the most significant result being the party’s first win in Corangamite since 1929. Darren Cheeseman built his win on especially solid swings from the Geelong booths and the large country centre of Colac, giving him enough fat to survive a surprisingly strong Liberal performance on postal votes.

Corangamite booth results with numbers varying in size to reflect number of votes cast. Click on image to toggle between two-party and swing figures.

Gavan O’Connor failed to make much of an impression in his bid to hold Corio as an independent, although his 12.7 per cent share of the vote earned him a handy $22,115 in public funding. His candidacy was probably responsible for a relatively mild 3.3 per cent swing to Labor. Interestingly, booths in Geelong’s industrial northern suburbs produced above-average votes for O’Connor and, in a number of cases, two-party swings to the Liberals. To the north of Melbourne, Labor achieved an evenly spread swing across Bendigo (5.2 per cent), Ballarat (5.9 per cent) and McEwen (6.4 per cent). This moved the first two from marginal to technically safe Labor status, but famously fell 12 votes short of delivering them Fran Bailey’s seat of McEwen. The swing against Bailey peaked at South Morang (10.6 per cent) and Wallan Wallan (9.1 per cent), but there was no clearly discernible pattern to its distribution.

With one exception, electorates in the west and north of Victoria saw big swings to Labor in regional cities and little movement in rural and small town booths. Sophie Mirabella suffered an especially strong 7.1 per cent swing in Indi, which peaked at over 10 per cent in Wangaratta and was only slightly smaller in Wodonga. Swings were considerably more modest at Mansfield, Euroa and Nagambie in the south of the electorate. Similarly, two large booths in Mildura produced double-digit swings to Labor that were not reflected elsewhere in Mallee, which swung 3.5 per cent overall. In Wannon, Liberal veteran David Hawker also suffered double-digit swings in a number of booths in the large coastal towns of Warrnambool and Portland, whereas there was little movement from most rural booths. However, this trend was bucked in Murray: booths in the dominant town of Shepparton produced swings consistent with the divisional total of 5.8 per cent, although there were some interesting exceptions in the city’s outskirts and its satellite town of Mooropna.

UPDATE: Great fun to be had in retrospectively comparing assessments of Gavan O’Connor’s prospects from Glenn Milne of The Australian and Jason Koutsoukis of The Age.

In a probably vain effort to maintain order around here, I will henceforth be running separate threads for discussion of the US presidential campaign. Here’s the first.

In Melbourne as elsewhere, the November 24 election produced a pattern of strong swings in outer suburbs and weak ones nearer the city, which cut across the partisan divide. The swing against veteran Liberal moderate Petro Georgiou in blue-ribbon Kooyong was just 0.05 per cent, while Peter Costello faced a similarly mild 1.7 per cent shift in neighbouring Higgins. This pattern carried over to the conservative dead zone of Melbourne, which swung only 1.1 per cent in Labor-versus-Liberal terms. However, the real story here was Greens candidate Adam Bandt’s success in edging out the Liberal candidate to take second place. The Greens’ primary vote was up 3.8 per cent to 22.8 per cent, 0.7 per cent behind the Liberals. This gap was bridged after distribution of minor party preferences, with Bandt leading the Liberal candidate 21,996 (25.1 per cent) to 21,405 (24.4 per cent) at the second last exclusion. Liberal preferences then took Bandt to within 4.7 per cent of victory, producing the first ever “Labor versus Greens” two-party result in a federal seat at a general election. This is the first time Melbourne has met the AEC’s definition of a marginal seat (6 per cent or less) since 1904.

Beyond the swing-resistant inner core of Melbourne, Kooyong and Higgins lay a band of seats separating it from the volatile outer suburbs. Batman followed the broader pattern of mild swings of around 4 per cent in inner suburban Northcote, and heavier ones of 6 per cent to 7 per cent at Preston and Reservoir further to the north. The Greens’ vote was up 3.2 per cent to 17.2 per cent, a potentially bridgeable 3.4 per cent behind the Liberals. Jagajaga, Chisholm and Menzies produced near identical swings of 4.6 per cent to 4.7 per cent, respectively staying safe for Labor’s Jenny Macklin and Anna Burke and the Liberals’ Kevin Andrews. On the bayside, Melbourne Ports produced a relatively gentle 3.4 per cent swing which was nonetheless the biggest movement in the electorate since 1993, while its safe Liberal neighbour Goldstein swung 4.0 per cent.

The two biggest swings were in the Melbourne area were in the outer suburban suburbs of Calwell in the north and Holt in the south-east. Calwell topped the table at 11.1 per cent, with swings topping 15 per cent at Craigieburn on the outermost urban fringe. The 10.1 per cent swing in Holt was most pronounced in the south, peaking with a mighty 17.5 per cent swing at the electorate’s largest booth of Narre Warren South. Swings in the north were in the range of 5 per cent to 9 per cent. In what might be regarded as the defining booth result of the election, a 10.96 per cent swing to Labor was recorded at Kath and Kim’s home of Fountain Gate.

Labor added some fat to a number of margins in traditionally safe south-eastern seats that were cut uncomfortably fine in 2004. After previous member Ann Corcoran suffered an unexpectedly close shave in 2004, newcomer Mark Dreyfus boosted the Labor margin from 1.5 per cent to 7.7 per cent in Isaacs, which produced heavier swings in the inland suburbs of Keysborough and Carrum Downs than along the coast. Immediately to the north, Simon Crean increased his margin from 7.4 per cent to 13.0 per cent in Hotham, with particularly strong swings recorded in Springvale. In Bruce the swing to Labor was an evenly distributed 4.8 per cent, increasing Alan Griffin’s margin to 8.3 per cent.

Liberal seats in the eastern suburbs mostly followed the trend of their Labor-held neighbours. Only in the case of Deakin was the swing enough for a seat to change hands, Labor winning the seat for only the second time since its creation in 1937. Their candidate Mike Symon picked up 5.7 per cent on the primary vote and 6.4 per cent on two-party preferred to prevail with a margin of 1.4 per cent, ending the 11-year parliamentary career of Liberal member Phil Barresi. Labor achieved an identical swing further afield in McEwen, which was famously 12 votes short of what was needed to unseat Fran Bailey. The swing peaked at South Morang (10.6 per cent) and Wallan Wallan (9.1 per cent), but there was no clearly discernible pattern to its distribution. Labor’s other disappointment was a 5.3 per cent swing in La Trobe that fell 0.5 per cent short of delivering them the seat. The Dandenong Ranges formed a rough dividing line between suburbs on the city side where the swing was in the order of what Labor required, and the hill suburbs and surrounding small towns where it fell just short at around 4 per cent.

On safer ground for the Liberals, Bruce Billson’s seat of Dunkley returned to the marginal zone with a swing of 4.2 per cent that was felt more heavily in Frankston than Mornington and Mount Eliza. In outer suburban and semi-rural Casey, Tony Smith suffered a 5.4 per cent swing that was higher in suburban Croydon and Kilsyth than in the satellite towns of Monbulk and Woori Yallock. It is interesting to note a particularly sharp 8.1 per cent swing in Aston, which memorably gave the Liberals a bigger margin in 2004 than Kooyong. Any thoughts that this might have marked a long-term realignment can now be laid to rest, as the respective margins are now 5.0 per cent and 9.5 per cent.

Labor’s safe seats in the west and north of Melbourne produced remarkably consistent swings of between 5.5 per cent to 6.7 per cent, excepting the aforementioned Calwell. Wills followed the pattern of neighbouring Batman in producing smaller swings of around 4 per cent at Brunswick at the southern end nearer the city, increasing to around 7 per cent at Glenroy in the north. Bill Shorten’s candidacy appeared not to make much difference one way or the other in Maribyrnong, which swung to Labor by a locally typical 5.8 per cent which was evenly distributed through the electorate. There was similarly consistency in the swings in Gellibrand (6.5 per cent), Gorton (6.3 per cent) and Lalor (6.7 per cent).