I’ve previously posted a number of times before about the dataset I have compiled of election results since 2004, adjusted for 2025 electoral boundaries.
Most recently, prior to the election, I specifically did a comparison of the 2007 and 2022 electoral map: two elections with similar two-party-preferred votes and a similar number of seats with a two-party-preferred majority for each major party.
Unfortunately there is no clear comparison for the 2025 election: it is the best result for Labor since 1943, with Labor polling a two-party-preferred majority in 100 seats. I also wouldn’t rule out the possibility they could take the lead in Bradfield too, but it looks like they fell just short.
So for this blog post I am going to focus mostly on how each seat is ranked, from 1 to 150, on the two-party-preferred vote, and how that rank has changed over time. I will consistently use the rank of 1 to refer to Labor’s best seat, and 150 for the Coalition’s best seat.
When the country overall swings by over 3% towards Labor, you’d expect most seats to swing towards Labor and that is true. 123 seats swung towards Labor, with just 28 seats swinging towards the Coalition.
I will finish this post with a map. If you click on a seat on the map, you can see my estimate of the 2PP and the rank in each seat at the last eight elections.
I will start by comparing ranks in 2022 to 2025. This first chart shows the ten seats that shifted the most towards the Coalition and the ten seats which shifted the most towards Labor.
Bendigo shifted the most, going from Labor’s 37th safest seat to their 95th safest seat. Interestingly the neighbouring seat of Ballarat had the third biggest shift towards the Coalition, but drew a lot less attention, as it is still clearly on the Labor side of the chart.
Braddon shifted the most towards Labor, going from the Coalition’s 31st-safest seat until it was on Labor’s half of the chart.
It’s also worth noting Fowler, which has moved 44 places towards Labor. This puts Fowler back into its traditional place on the rankings. The seat was ranked 7th-15th at all but one election from 2004 to 2019, but the 2022 campaign pushed the seat towards the centre. While Dai Le managed to retain her margin, the Labor 2PP has become much stronger. I assume we’ll find that the proportion of Le’s voters who preferenced Labor over Liberal will have increased significantly, but that’s a question for next week.
For this next chart I show seats that flipped over the line that separates the 75 best Labor seats from the 75 best Coalition seats.
The three northernmost Tasmanian seats all moved to the Labor side of the chart. Solomon and Lingiari swapped sides. Outer suburban seats like Bullwinkel, Pearce, Whitlam and Werriwa have all moved to the Coalition’s top 75, although Robertson has moved the other way.
I have also looked at how seats have shifted over the longer time frame from 2004 to 2025.
A lot of these seats had already shifted by 2022, and haven’t changed as much in 2025. But Flynn has continued to shift towards the Coalition, moving from #104 in 2022 to #135 in 2025. Whitlam has also moved 26 ranks towards the Coalition in 2025, on top of moving 14 ranks over the previous 18 years. Moreton has also continued to shift towards Labor, going from #48 in 2022 to #27 in 2025.
This fourth chart shows seats that have moved from the Labor side to the Coalition side and vice versa over 21 years. There are ten seats that have moved in each direction.
My attention was drawn to the neighbouring seats of Werriwa and Macarthur, which have moved by 58 and 34 ranks respectively, but in opposite directions. This final chart shows the ranks of these two seats since 2004.
Werriwa was one of Labor’s safest seats in 2004 and 2007. Mark Latham left the seat in 2005. The seat bounced around but since 2016 it has steadily become less safe, until in 2025 it was Labor’s 76th best seat. Macarthur, on its current boundaries, has always leant towards Labor, but was usually one of Labor’s less safe seats in the 2004-2013 period.
Mike Freelander won Macarthur in 2016, and the seat became a much stronger Labor seat that year.
That’s it for today. This map shows three layers: the change in rank from 2022 to 2025, the change in rank from 2004 to 2025, and the biggest net change in ranks over the 21-year period.
Interesting analysis, thanks Ben! I think this ranking system is a good basis for the Coalition to strategise around the roughly 10 to 15 seats it needs to gain in 2027/28 in order to set itself up for victory in 2030/31.
Though I’m not sure if the Coalition needs to mainly focus on the seats that have been shifting blue or the ones that have been shifting red. Probably a bit of both as they will need to start build a voting coalition that is more than just regional electorates, whether that is through outer suburban, multicultural or upper middle class electorates.
Obviously all of the recent Labor gains are up for grabs as anything that swings one way can always swing back. But as per your charts, it now looks like Bendigo, Bullwinkel and Solomon are new key targets for the Coalition, followed by Pearce and Whitlam. Blair and McEwen also remain as solid options.
I’m wondering how much the analysis would change if using 2007 as the baseline as that was the last “Labor Landslide” even though the seat count way quite different. It’s quite impressive that the party has gotten to 94 seats despite having lost Capricornia, Dawson, Flynn, Lindsay, Longman and Page since the 2007 election. Herbert and La Trobe have also briefly been held in the intervening period.
There are seats where a party does very well and picks up or gets a swing and then entrenches thenselves for several more election, often times with the same MP reelected again.
The working-class backlash brought about by the Adani Carmichael mine issue damaged Labor’s brand in regional QLD, especially in Capricornia which Labor lost in 2013 and hasn’t regained since. Capricornia was mostly a Labor seat before the LNP landslide of 2013.
On the flipside, Labor gained Kingston and Blair in 2007 and the exact same MPs hold them to this day. Greenway is another example of where the Labor brand is entrenched but it’s likely to do with changing socio-economic and demographic factors and not just Michelle Rowland’s personal vote.
I note that in teal-held seats, Labor’s 2PP has significantly increased over 2 elections.
interesting Analysis
Some points
1. When looking at a long term trend from 2004,McEwen is fairly neutral when Libs had a sitting MP they overperformed the Victorian state vote. Also increasing urbanisation has helped Labor
2. Fowler had a big notional swing back to Labor on Ben’s post prior to election it looked like a blue trending seat.
3. Bruce and Hawke are red trending compared to 2004-2025 while they were blue trending 2007-2022.
4. I think Moreton is now a safeish Labor seat almost like Lilley which Libs can only win at huge landslide in the case of Lilley greater than 1996 and Moreton a 1996 level.
5. Parramatta is Red trending 2004-2025 while was blue trending 1996
6. La Trobe is actually not the same seat that Labor won in 2010. Labor would not have won in 2010 on current boundaries but Labor actually overperformed the 2010 result. It is probably the only seat in Victoria that Labor did not win in 2025 that Labor could win in the future.
7. @ Votante, Makin is another seat that was gained in 2007 that is still held by the same MP
8. Bullwinkle should really be a Lib seat there is no excuse for Libs to have failed to win it in 2025
9.Solomon is a traditional marginal seat
I think Liberal targets as part of the road to recovery include:
1. Outer suburban seats that weren’t held 2013 to 2022 e.g. McEwan, Bullwinkel
2. Provincial or outer suburban seats that flipped, including those with margins that could be inflated due to one-off, state-based factors e.g. Bass/Braddon (though an 8% swing is a big ask), outer suburban Brisbane seats and possibly Hughes as well.
There are some 2022/2025 seat changes that are realigning and trending to Labor long-term. This would make it harder for the Liberals to regain them especially if the sitting Labor MPs perform well and demographic trends continue e.g. Bennelong, Reid, Bonner, Sturt, Tangney.
@ Votante, Just a few points
1.Bullinkinkle would have been Liberal held had it existed in 2004 until 2022.In essence it should be a Liberal seat. It either replaces Stirling (WA) or Higgins (VIC) both of which are tend to Liberal seats so no real numerical advanatage for Libs by winning it back.
2. McEwan i agree it can be won by Libs
3. I think it is too early to say whether Bennelong, Reid, Bonner, Sturt and Tangney will long term trending Labor. Between 2004-2019 Reid and Bonner Trended Liberal especially Reid. https://www.tallyroom.com.au/47432
4. Also even if they win Dobell, Eden Monaro, Solomon, Lyons all 2016 loses that will not be enough to make up for the seats in Point 3. Cowan is also a 2016 loss but compared to Tangney is more CALD, Muslim and less affluent.
Essentially there’s no way back to power for the Libs if they ignore the middle ring CALD seats. Dutton is basically one of the worst people to appeal to for those kinds of seats.
Bullwinkel surely would’ve been a Liberal seat if it had existed before 2022. It was only notional Labor due to Labor’s overperformance in WA.
Bennelong, Reid, Tangney and Sturt had above-average swings to Labor in 2022 and 2025. The first 2 seats had swings to Labor in 2019 when the country moved away from Labor. They are also quite CALD and have large numbers of white-collar professionals.
Eden-Monaro and Cowan in theory are winnable for the Libs but their local members seem to be popular and have built up pretty big margins.
@ Votante
Important to remember Cowan and Burt are the most CALD seats in WA and more CALD than Sturt and Tangney i cant see a scenario where Libs fail to win Tangney but win Cowan. As Perth is a smaller city seats are more diverse demographically but Cowan and Burt have Red wall suburbs