The scale of Labor’s victory in 2021 makes it hard to know what seats are likely to be the most interesting seats at the 2025 Western Australian election, or the shape of the likely result.
Recent polling points to a Labor two-party-preferred vote of 55-56% – five polls conducted between July 2024 and February 2025 all gave a result in that range.
Conveniently, that is pretty much exactly the Labor 2PP achieved at the 2017 state election. Labor polled 55.5% that year, and then gained a swing of 14.1% for a result of 69.6% in 2021.
If these polls are correct, and Labor again polls a 2PP similar to the 2017 result, you couldn’t necessarily assume that each seat will swing back by the same degree they swung in 2021. Indeed some seats swung harder than others in 2021, and it would also not be reasonable to expect that the relative position of each seat will stay the same. I expect some of the seats that had bigger swings will have bigger swings back to the Liberal or National parties, while others may remain more strongly pro-Labor than they were prior to 2021.
So for today’s post I’m going to explore some various scenarios for how the state might shake out for a result in that range.
For this blog post I am looking at four different scenarios:
- The actual 2017 results
- The 2017 results adjusted for the 2021 electoral boundaries
- The 2021 results, with the Labor 2PP reduced by 14.1% in every seat.
- The 2021 results adjusted for the 2025 electoral boundaries, with the Labor 2PP reduced by 14.1% in every seat.
Western Australia’s elections still remain very dominated by classic seats, with the vast majority of contests in 2017 and 2021 being races between Labor and either Liberal or Nationals. The only exceptions were Moore and Roe in 2017, which were Liberal vs Nationals races; Baldivis in 2017, which was Labor vs independent; and Fremantle in 2021, which was Labor vs Greens. I’ve excluded these races from the charts, but they are included in the maps.
I also haven’t tried to treat Liberal and Nationals. In some places, the expected swing back from Labor will benefit the Liberal Party, in other places the Nationals. But that’s another thing to calculate. So I am treating them as a single bloc. I can’t say “Coalition” in this context so where I am sicking of saying “Liberal and Nationals” I’ll say “conservative” with a small “c”.
Firstly, this chart shows the 2CP in all classic seats for all four scenarios, based on a Labor 2PP of 55.5%.
In three of these four scenarios, the Liberals and Nationals would win 18 out of 59 seats, which is what actually happened in 2017, while Labor would win the other 41. In the case of 2021 results (minus 14.1%), the conservative parties would have 20 seats.
In each scenario, the extra 2PP swing that the conservative parties would have needed to win their thirtieth seat (and thus a majority would have been):
- 5.8% – 2017 results
- 7.9% – 2017 results, 2021 boundaries
- 9.3% – 2021 results
- 8.5% – 2021 results, 2025 boundaries
Bearing in mind that Labor actually won a 5.5% margin statewide in this scenario, it suggests the electoral maps since 2021 have been substantially favourable to Labor.
That above chart shows that the conservative parties in 2017 won very few seats by a slim margin, while Labor won a lot more. This was smoothed out a bit by the redistribution, but you can still see it at the top of the pre-2021 pendulum. There are a lot more Labor seats than Liberal seats with margins under 3.5%.
Prior to the 2021 election, Labor’s seats were very evenly distributed along the pendulum. Using the post-redistribution estimates, Labor held 6-7 seats for every 4% of margin from 0% to 24%.
Things changed in 2021. Labor won eight seats with margins of 10-14% (which thus become marginal conservative seats in our scenario), but otherwise their seats were distributed in a normal shape with a peak around 76-80%.
If there was to be a uniform swing back on the 2021 boundaries, you’d expect the Liberals and Nationals to win a lot of seats narrowly and Labor wouldn’t win many seats by slim margins.
This was also flattened out a bit by the most recent redistribution. This has resulted in more seats on either side of that 14.1% swing line, but it’s also increased the peak around 76-80%.
Finally, this map shows where each seat fell under each scenario. If the polls are right in 2025 it will be interesting to see how different the 2025 map will look compared to 2017, and where the swings will be biggest.
The changes between the first and second scenario (and the third and fourth) mostly reflect redistribution changes, but the change between the second and third are all about differential swings in 2021.
Labor tended to do relatively better across Perth in 2021 than in 2017, while they did less well in places like Bunbury, Collie-Preston and Albany.
Indeed if you grab every seat that was classic in both 2017 and 2021 and compare the relative rank of each seat in both those years, some seats stick out as changing their position quite dramatically.
Labor’s swing in Southern River was 11% more than the statewide swing, and it was also particularly large in Joondalup, Burns Beach and Wanneroo. Southern River is in the outer south-east of Perth, while the other three are in the outer northern suburbs.
On the other hand seats like Albany, Collie-Preston, Darling Range, Kimberley, Mandurah and Willagee had much smaller swings. The swings there were around 7-10% to Labor, compared to 14.1% statewide. This could suggest that these seats won’t swing back as hard as other seats, but could also suggest that the enormous Labor landslide covered up a relative weakening of support here. It’s worth noting that three of these seats are regional.
We don’t have the same kind of seat-specific or region-specific polling in WA that we’re getting for the federal election, so there’s a bit of guesswork here, but there is a history of seats not falling in the order that you would expect.
The 2011 NSW state election was a landslide victory for the Coalition, and there was an expected swing back in 2015. Yet some super-marginal Coalition seats like East Hills and Monaro did not change hands. The expected swing away from Labor in WA in 2025 is about twice the size of the swing in NSW in 2015, so it’s unlikely we’ll see super-marginal seats still in Labor hands, but there could be some unusual results further up the pendulum.
I suspect Labor will retain most of their regional seats. Country people knowing Labor will be reelected want a voice in govt.
The urban seats won’t be just 2017/2025 boundaries… they will vary.
Seats like South Perth Scarborough may be retained.
The Darling range/ Kalamunda could be lost..
Suspect natural liberal urban seats like Nedlands Church lands will be lost to Labor but some wins for teals possible
@mick all the seats you just mentioned will be lost. labor will lose most of the country seats. they will lose Kalgoorlie, Warren blackwood, murray-wellington, dawesville, geraldton the only hope they have in albany is th dumb position of the local liberal candidate
If I had time I’d want to estimate based on something giving the same overall average:
* Incumbency (higher swing for a retiring incumbent)
* Sophomore surges (lower swing for candidates seeking precisely a 2nd term)
* Both the above being larger factors in regional areas, where a local favourite can more easily become entrenched.
* However the overall swing against Labor being higher in regional areas (this may precisely cancel out the above?)
With all of those combined, I expect a surprisingly high seat haul for Labor, Labor hanging on to some of the unprecedented 2021 gains and losing some 2017 gains or even 2013 retains.
nats may even be close in pilbara and kimberley
Darren Spackman might be an upset in Kimberley, it’s a similar circumstance to how Pauline Hanson was elected in Oxley in 1996.
This will be a fascinating election. More correctly, the difference in voting in the state and federal election in WA will be fascinating. There is a state government against no valid opposition. The Liberals should have done with Zempilas what QLD did with Newman. But views on the ground are that Labor will outperform in the State election compared to the Federal election given the premier of WA is essentially a pro-gas expansion conservative and west Australians don’t want two elections in a matter of weeks. WA is a sleeper state for the federal election and polls aren’t really picking it up given small samples.
Everyone now knows the liberals will not win in wa . People want a voice on govt..
So Labor will hold more than expected.
Re country areas. .. nor normally labor’s happy hunting ground. But when Labor wins as long as the sitting mp is good they are very hard to defeat. Look @ Mick Murray and Peter Watson before they retired .
It’ll be interesting to see if any seats fall that weren’t won by the Liberals in 2013 such as Albany.
@mick mick murray was in what is considered safe ALP seat due to the unionised owrkforce in collie which is the main part of that seat. peter watson same deal albany has been a fairly safe seat for labor and that has always been marginal. so i imagine peter watson is a good member however it will swing hard against labor this time. and i imagine the only thing saving rebecca stephens is the stupid homophobic and anti abortion views of the liberal candidate so i think the nats may get into the 2cp this time. they might win it from there but im gonna say marginal labor retain. shell be in trouble in 2029 though/
@np if it falls it wont be to the liberals
Check Mr Murray ‘s electoral history
Definitely personal vote here.
Still helped by the high Labor vote that has kept it voting Labor.i use to live in this seat
Collie with 70 % plus 2pp alp vote certainly helped .
Minus Mick Murray Labor would have lost 2 or 3 times since 200q.
A 14% swing back and a 55% ALP victory looks pretty similar to the 2022 fed election. So mapping those results onto state boundaries would also give you an idea. Would have Labor being competitive to win in South Perth and some of the state seats within Tangney.
Mainly the seats within Curtin, parts of Moore, parts of Tangney and the more conservative parts of the ritual and regional seats would be won by Libs or the Nats.
you have to wonder what made people change their mind from hilda Turnbull in 1996 to Mick Murray in 2001 to the tune of almost 10% given the same 2 candidates contested the 1993, 1996 and 2001 elections. its just a matter of when the libs/nats win this not if
You have to remember that the predecessor seats of Collie – especially when Collie went to coast, had the strangest results in that the seat had the highest Labor Booth in state and the Highest Liberal Booth in the state. This was around the 2000 – 2010. What other seat would have that feat. I always thought it was strange that the Nationals held Collie because there is not a lot of farming and Nationals territory there.
2001 was a change of govt election
Labor won Geraldton and Albany at the same election. In one of those the sitting lib was in trouble for some impropriety… don’t remember what it was about or whether he was guilty… Mr Prince ….Albany I think.
Almost politics 101. . It is difficult for Labor to win most country seats… but once they do… they tend to stay
Happens all the time in nsw at the state level.
John, One Nation preferences were a big part of that – they put all sitting MPs last on their HTV cards in 2001, no matter who it was. That helped Labor win Collie, Albany and Geraldton (and also lose Kalgoorlie). A lot of seats also had the conservative vote split three ways between Lib, Nat and ON.
Mick, it was both of them – Bob Bloffwitch in Geraldton and Kevin Prince in Albany. Also Doug Shave in Alfred Cove. Tory sleaze was a factor that year.
James: Collie (the seat) used to be further east – in 2001 it contained the shires of Boddington and Boyup Brook, which are Nat voting farming regions. The redistribution that created Collie-Wellington pulled it into the south-west, where the Libs traditionally get the conservative vote.
Bird
Thank you remembering back over 20 years ago can be hard some times
Reading down the pendulum…
Churchlands, Nedlands, Carine, Bateman, South Perth – the Libs just have to get these back. South Perth may be a bit of a roughie because of all the new high-rise flats there lately – that’ll be making it younger, but probably no less Lib-voting considering the money you need to live there. Churchlands is the vaguest of chances because the candidate is Zempilas.
Warren-Blackwood is in the same boat, but I wonder if the Libs could take that rather than the Nats. In 2008 this was a merger of a Lib and Nat seat, with the Lib retiring. Now the Nats lose Terry Redman’s personal vote, so the Nat-Lib margin might shrink. It may end up depending on Labor and Green preferences.
Scarborough and Riverton are decent chances below 15% for Labor. They get a sophomore swing in both seats, and Riverton tracks the statewide vote pretty closely since 1989 – it’s only in 2017 and 2021 that it’s picked up a Lib lean, thanks to the swing to Labor being heavier in the outer suburbs. (2017 was the first time since its creation that it’s been held by an opposition MP.) Scarborough is similar – Innaloo was a Labor seat on similar boundaries in the early 2000s.
Albany, Geraldton, Kalgoorlie: all regional, so they’re dependent on whether the new MP has gotten established. Kalgoorlie is an unclassifiable four-way mess thanks to the former Lib MP running as an independent.
Dawesville: I’m still tipping a Lib gain, but they get a double sophomore effect here (indeed, every seat below 15% except Kalamunda), so it’ll be close.
Darling Range was a seat Labor won in 2017 but then lost in a by-election, so it didn’t get the same huge swing in 2021 that other outer suburban seats did. It’ll probably get a smaller swing back to compensate, and then be one of the marginals that’ll decide the 2029 election.
Kalamunda has a retiring MP and is right on the expected swing, so it’s probably gone.