Queensland state redistribution drafts published – live

10:54 – I’m going to finish this liveblog now. Possibly to be revisited a bit more succinctly in coming days.

10:46 – Here are my estimates of primary votes.

10:39 – I have now finished my estimates of two-candidate-preferred margins. In twelve seats, a seat contains areas that had different 2CP combinations in 2024, so I’ve needed to make estimates by applying preference flows from the larger portion to the smaller.

The LNP has gained two extra seats, going from 52 to 54. Labor has lost one, while KAP has also lost one. Gaven becomes LNP seat, then Labor and KAP each lose an abolished seat, and Labor and LNP each gain a new seat.

On the old boundaries, a 1.9% two-party-preferred uniform swing would deprive the LNP of their majority. On the new boundaries, that swing would need to be 2.9%. That is a significant improvement for the LNP. There is no change in the uniform swing needed for Labor to win a majority (3.5%).

Looking at seats that have a margin under 4% either before or after the redistribution:

• LNP is 4.6% stronger in Gaven
• Labor is stronger in Aspley, Caloundra, Rockhampton, Redlands, Cook
• LNP is stronger in Pumicestone, Mirani, Beenleigh, Springwood, Capalaba, Cairns, Mulgrave, Nicklin/Nambour, Clayfield, Logan, Ipswich West
• Greens stronger in Maiwar/Indooroopilly and South Brisbane

It does seem like the LNP has done better in a majority of marginals: 12 for LNP, 5 for Labor, 2 for the Greens. The Gaven redrawing is particularly helpful for the government.

Margin estimates

10:20 – Just to summarise, I have found a number of seats where I think the Commission has been mistaken in matching the old seat to the new:

• Greenbank is the new seat, not Springfield. The old seat of Jordan has transferred much more substantially to Springfield than Greenbank.
• Toohey was abolished, not Stretton. The new seat of Eight Mile Plains overlaps substantially more with Stretton.
• Marsden is not a new name for Waterford. Rather the name of Woodridge was applied to the old seat of Waterford, and the old seat of Woodridge was renamed Marsden.

Once you factor in those changes, the proportion of voters moved has dropped to 14.37%.

6:42 – Hinchinbrook has also lost about half of its territory, expanding to take in half of the abolished seat of Hill, and losing areas closer to Townsville to its neighbours. Seats like Glass House, Gaven, Flinders, Oxenford and Annerley have 33-43% new residents.

6:40 – Greenbank is supposedly a new name for Jordan, but only 39.1% of Greenbank voters have come from Jordan. Eight Mile Plains is supposedly a new name for Toohey, but 57.3% of voters come from the abolished seat of Stretton. So I would argue that it is actually Toohey that that has been abolished, not Stretton. But perhaps I’m wrong? Let me know.

This lowers the number of voters moved to 14.85%.

6:31 – This tweak lowers the number of voters in a new seat down to 15.01%.

6:29 – As for individual seats, some seats have changed to an extreme degree:

• 81.9% of voters are new to “Marsden”, which the ECQ claims is a new name for Waterford. 62.5% of Marsden voters have actually come from Woodridge.
• Meanwhile just 37.5% of voters in the new Woodridge came from the old Woodridge. 67% have actually come from Waterford.

So I would argue that Woodridge is a new name for Waterford and Marsden is a new name for Woodridge.

6:24 – Thanks to William Bowe for sharing some data that I’ll need to calculate my margins, I estimate that 15.8% of people have been moved to a new seat. This assumes that each renamed seat is actually the successor to the seat that the report claims it is replacing.

3:56 – You can use this map to toggle old and new boundaries for the state, or use the ECQ map here. It’s quite a large file but it’s about as small as I could make it.

3:04 – The ECQ has published spatial boundary files as a geoJSON file. Unfortunately I’ll need some more time to convert it into the format I use, and to match it to the enrolment by SA1 to determine how many voters have been moved from old to new seats before calculating the redistribution.

2:57 – At first glance it does resemble the LNP submission.

2:56 – Unfortunately it looks like they have not published geospatial files which severely limits our ability to calculate seat margins. Blame the ECQ. Hopefully they will share them soon – otherwise I’ll have to draw the boundaries from scratch which will take weeks.

2:55 – Hill and Stretton have been proposed to be abolished, and two new seats of Caboolture and Springfield are proposed to be created.

19 seats have been renamed:

• Bancroft as Deception Bay
• Bonney as Labrador
• Bundamba as Redbank
• Chatsworth as Carindale
• Coomera as Pimpama
• Cooper as Ashgrove
• Jordan as Greenbank
• Macalister as Beenleigh
• Maiwar as Indooroopilly
• McConnel as Brisbane Central
• Miller as Annerley
• Nicklin as Nambour
• Ninderry as Coolum
• Oodgeroo as Cleveland
• Scenic Rim as Beaudesert
• Theodore as Oxenford
• Toohey as Eight Mile Plains
• Traeger as Flinders
• Waterford as Marsden

Quite a few of these are reversing renamings we saw in 2016.

2:50 – The redistribution has been published! And I have 20 minutes to look at before going out with family responsibilities.

2:04 – The draft boundaries for the Queensland state redistribution are apparently due to be published some time this afternoon. I expect the maps to be published on this website.

Unfortunately this is a bit of an inconvenient time for myself – I will attempt to analyse them as quickly as possible but a full analysis may need to wait for tonight or tomorrow.

This is the first redistribution of Queensland state electorates for a decade, with the current boundaries having been used at the 2017, 2020 and 2024 elections.

My plan for this afternoon and evening is to broadly describe proposed changes, and then calculate the partisan impact of the changes as well as the volume of voters who have been moved to a new seat.

Liked it? Take a second to support the Tally Room on Patreon!