Now that the Parliament has officially opened for its first sitting after the federal election, we’re off to the races with redrawing the electoral map for the next federal election, expected in 2028. No rest for the wicked.
The redistributions in this parliamentary term are expected to be much less dramatic than those seen in the previous term. It seems unlikely that any state will see a change in their entitlement of seats in the House of Representatives. If so, the only redistributions will be the four triggered due to seven years passing since the previous redistribution. Conveniently, this will mean that the four states and territories that did not redraw their boundaries will get a new map in this term, while the other four will be left alone.
In this post I will run through the timing of these processes, and the data we now have to give us a hint about how these redistributions will play out.
Under Australian law, the federal electoral boundaries for a state or territory will be redrawn for one of three reasons:
- A change in that state’s entitlement to seats in the House of Representatives
- Seven years has passed since the conclusion of the last redistribution.
- The number of electors in at least one third of divisions deviates from the average enrolment by more than 10% for at least two months.
The third criteria has never been triggered, thanks to the AEC’s strict rules on drawing boundaries (more on that later).
Under the second rule, there are four jurisdictions due to commence redistributions this year. Queensland, South Australia, Tasmania and the Australian Capital Territory all completed their last redistributions in 2018. The seven-year mark was reached earlier this year for Queensland and Tasmania, but those redistributions were deferred until after the federal election. South Australia and the ACT have both reached the seven year mark this month. So they should all start very soon.
The other question is whether the entitlement of seats per state will change at all. This calculation will be conducted one year after the first sitting of Parliament (July 2026) using the latest ABS population estimates at the time, which will be the December 2025 data. This data is published every three months, so we can track the trajectory.
This chart shows the quarterly ABS population estimates from March 2020 to December 2024 (the latest currently available). It shows how much each state’s population is above or below their current seat count.
There is no state in any danger of currently losing a seat. New South Wales is on a downward trajectory that could cost the most populous state its 46th seat prior to the 2031 election, but not in the next year. South Australia is also on a downward trend.
Queensland is getting relatively close to gaining a 31st seat, but it doesn’t look like it will be enough. Queensland currently has 30.36 quotas of population, so needs another 0.14 quotas. It has taken over two years for Queensland’s population to grow by 0.14 quotas.
If there is any doubt about whether Queensland’s seat entitlement will change, the AEC has the option of deferring Queensland’s redistribution until the decision is clear. I don’t think Queensland will gain a 31st seat, but I don’t know if that deferral will happen.
All the same, that means that it’s just these four jurisdictions (most of them relatively small ones) being redrawn.
While in 2022-25 there were redistributions affecting 102 electorates (with 100 in states where the seat count changed), in this parliament that number will likely be just 48 (and none where the state’s count changes).
This parliamentary term would thus be only the second term since at least the 1993-96 team where there was no change to the seat entitlement for a state or territory.
This should mean that the scale of the changes will be reduced. It will also allow me to prepare my election guides for the unredistributed two thirds of the country much earlier than for the redistributed parts.
So, what is likely to happen in these four jurisdictions? For this next section I will post the current (June 2025) enrolment statistics for seats in each state.
Federal redistributions require divisions to be drawn within 10% of the average as of the current day, but more importantly they must be drawn within 3.5% of the average based on projected enrolments three-and-a-half years in the future. We don’t yet have projections, so I can only do analysis on current enrolment data.
Let’s start with Queensland, with its 30 electorates. This should not be confused with the redrawing of Queensland’s 93 state electorates, which has just recently commenced.
This map shows how much each seat deviates from the average: red for under-average seats, blue for over-average.
The most distinctive feature is a band of seats around South-East Queensland that are well over quota: Wright, Blair, Longman and Fisher are all over 10% over quota. Fadden and Forde, which border each other on the edge of Brisbane and the Gold Coast, are not far behind.
To simplify the story, this table shows how much each region of the state deviates from the average.
The 20 seats of south-east Queensland are collectively about one third of a seat over quota, but this is not evenly distributed. Indeed six seats in the outer south-east (which includes Ipswich, the Sunshine Coast and some other areas) is more than half a seat over. The Gold Coast is roughly in line, while Brisbane is almost a third of a seat under population.
This points towards Brisbane seats having to expand outwards to absorb that surplus population in those outer suburban seats. In the Brisbane area proper, the seat of Brisbane is slightly over quota, and Petrie is substantially over quota, but every other seat is below average. Moreton is particularly below average.
Outside of the south-east, just Hinkler has an above-average population. No seat has a particularly large shortfall, but it just adds up over the ten seats.
The same outer-suburban trend can be seen in South Australia.
Spence, in outer northern Adelaide, is 6.6% over quota, while Mayo is 6.3% over.
Hindmarsh is pretty spot on the average, while the other seven seats are all under quota. Makin is particularly under quota, by 5.8%.
The seven seats in urban Adelaide bounded by Spence and Mayo are pretty much exactly 10% under quota, so they should be able to absorb most of the surplus population in Mayo and Spence, although they are likely to be projected to grow faster than the two regional seats of Grey and Barker.
Tasmania’s new electoral map will apply to state and federal elections, following the recent Tasmanian state election.
The main story in Tasmania is the significant population shortfall in Clark, where enrolment is 9.9% below the average.
Braddon is slightly over quota and Bass is slightly under, but the two cancel each other out. Franklin is slightly over quota, but Lyons is significantly over quota.
It seems likely that this imbalance can be solved by Clark expanding further into the Hobart suburbs, in particular taking in parts of the southern end of Lyons.
The options get simpler for states with a smaller number of electorates. The ACT has just three seats.
The southern electorate of Bean is substantially over quota. Canberra has the biggest shortfall of voters, while Fenner is also slightly under. The exact changes would be dictated by the projections, but it’s likely Canberra will need to expand towards Woden to restore the balance.
John, you are correct. In fact, the redistributions have formally commenced today (except Queensland, where the commissioners felt a further deferral until July next year is best in case the state is entitled to an extra seat in reapportionment).
THe AEC has announced that federal redistributions have begun in SA, TAS and ACT, while the federal redistribution for QLD has been deferred until the second half of 2026, after the number of electorates each state and territory is entitled to is determined in July 2026. https://www.aec.gov.au/media/2025/08-12.htm
Suggests will open on 8 October, when enrolment projections will be made available.
Yea i saw that coming. They acted on the side of caustin but i dont thin qld will be entitlrd to another seat. Its a shame really because it would of been so easy to do be ause qlds state redistribution numbers made it so easy. Just use 3 state seats for every federal seat.
Caustin would be a great name for a new federal division.
Lets assume we dont know who that is
Queensland will probably hit its 31st seat around late 2026/early 2027. So about a year after the determination is made in my opinion so probably around the time they finish its redistribution. WA will probably make it too the 17th seat around the time of the next determination in 2029. Though I think NSW will just survive the determination that same year so instead of the 2031 election I’d say by the 2034 election. But by that time it’s 7 years is already up but it would then be deferred until the next determination. Otherwise Vic, Nsw and wa are all scheduled to be after the 2031 election. The act, SA and Tasmania will then be deferred again in 2033 due to being within 12 months of the 34 election. Assuming the govt of the day doesn’t go early. By that time though we may have an expansion of the parliament. Which is overdue.
@John I think any expansion of Parliament will be into the mid-2030’s at least, and even then it would have to be bipartisan. The big problem would be the Senate increase.
@John I think it’s going to be a lot closer than that. Just repeating my thoughts from the wrong topic because I got distracted. 🙂 And I’ve put the boring bit of the Act underneath because even I was starting to doze off reading it…
So the seat count as as December 2024 was 30.36. We have approximately 19 months of growth before the determinations start – roughly 0.11 quotas over that period assuming linear growth. That puts it at 30.47 seats. That’s a bit hard to determine at the moment because we’re relying on growth estimates and not actual data.
What I’m looking at now is the housing approvals paperwork. BCC is already concerned because the low and low-medium density housing applications have cratered for 2025 (mostly due to BCC’s own internal processes) and the State Government is starting to amend approved Priority Development Areas (PDAs) to make it more viable for developers to build now on sites held for landbanking. Interest rate changes, and anticipated changes, are also moving the market.
HOWEVER, and it’s a big however, the AEC is dependent on the Australian Bureau of Statistics to calculate the determination. They can ask, but the ABS might be a little busy in the middle of the 2026 Census. The advantage of the ABS delaying it’s response until the first results are through means the numbers will be as accurate as they can be, and the ABS can simple shoot the raw population numbers through to the AEC ahead of the release of the full Census. Just because they ask the Australian Statistician for the numbers it doesn’t mean he has to reply immediately. Normally the process is to give an approximate value based on extrapolating the quarterly figures and tinkering with average growth rates. When you’re miles away from affecting the quota then no-one checks at the next redistribution if you got it right.
Under s.59 of the Electoral Act, “A redistribution of a state or territory may be deferred if:
the redistribution is required due to the time requirements in the legislation (the last redistribution for that state or territory was determined seven years ago), and
the three-person Electoral Commission’s formal direction to start the redistribution is due within 13 months after the first sitting day of the House of Representatives, and
the Electoral Commissioner’s determination of the number of members of the House of Representatives has not yet been made. During the 13th month of a new Parliament, the Electoral Commissioner is required to determine the number of members of the House of Representatives for each state and territory using population data supplied by the Australian Statistician and in accordance with procedures set out by the Electoral Act, and
the three-person Electoral Commission is of the opinion the Electoral Commissioner’s next determination of the number of members of the House of Representatives for each state and territory may or will change the number of members in the House of Representatives for the specific state or territory.
If a redistribution was deferred due to this scenario, it is then required to start after the Electoral Commissioner’s determination has been published. This determination may or may not cause a change to the number of members for the state or territory at the next general election.”
Again your wrong. The growth will be 12 months as it will be based on the numbers released in june 2026. Which will be the numbers from the december 2025 quarter. Which is 12 monyhs from the current numbers. Even if the ABS delays their response the AEC will simply run off the most recently provided numbers.
According to the AEC website is slated to be determijed on July 23rd 2026.
@Darth Vader So you’re absolutely positive the AEC won’t announce a third deferral following their previous deferrals on 28th March 2025 and 12th August 2025. Also the AEC is REQUIRED to ask the ABS for their numbers – that’s part of the process. They don’t just wing it – specifically “using population data supplied by the Australian Statistician and in accordance with procedures set out by the Electoral Act”.
@ mark, Darth Vader is correct as to the figure that the AEC will be using
Interesting the AEC monthly figures and the VEC monthly figures were issues yesterday (for the first time on the same day for long time) and there is 4,000 voters difference, but both showing a decrease in the number of voters in Victoria over the last 2 months.
However the population figures that the AEC uses will be the ones determined for December this year and released in June next year.
I don’t believe that Queensland will get an extra seat this time, while the number of people moving to Queensland is massive, QLD also has the highest death rate which impact on its growth rate. This is not a problem that WA faces, which is why it will grow quicker than other states
The AEC was able to defer the start of the QLD redistributions for good reasons. They cannot defer the determination of the entitlement. That happens on a specific date using specific data. Once that is done, there will be no further reason for a deferral. I don’t see why the AEC would want to defer it anyway.
From the release:
“Regardless of whether this entitlement changes, a federal redistribution for Queensland must happen in the second half of 2026, as it has been more than seven years since the state’s previous federal redistribution.”
Should QLD get a 31st seat, it will be interesting to see where the new seat will go. Based on 31 seats and the 31 July enrolments – the combined Petrie+Longman+Fisher+Fairfax are 53% over quota and the combined Forde+Fadden+Wright+Blair are 57% over so it is not immediately obvious if it will be north or south of Brisbane
Redistributed theres also a 29% defeceit in bonner moreton rankin and oxley. It would result in blair losig somerset. The new seat would be north of the river. Bwver there likely wont be a new seat thus time around so those surpluses wil be redistributed this time around
I have returned to oz safely. Whenare we expecting these to kixk off
@Darth Vader, according to the press release, public suggestions start 8 October. Obviously, the detailed projections will be available from that date. They sometimes give the current elector quota earlier, but I don’t think they’ve made any announcement about it.
I’ve re-written my introduction, methodology and division names commentary for the new reports and started writing my analysis of existing divisions.
Now I just need to play around with producing maps in QGIS and having them correctly attributed for publishing so I’m good to go come October.
Hi everyone, I have a question for those of you who have dived deep into specific boundaries during the last round of redistributions, for a blog post/submission I’m working on.
Do you have examples of odd and non-intuitive boundaries drawn during the 2022-25 redistribution by the commission which only make sense in the context of minimising the number of total voters changing seats. I’m only looking for new boundaries.
The one example I have is Kingsford Smith expanding into the former Rockdale council area. It would have significantly reduced the number of voters moved compared to a more intuitive boundary that would have caused greater knock-on effects on Sydney, Grayndler and possibly even Wentworth.
Ben, the other boundary drawn which does not seem to reflect on community of interest is the appendage of Hughes into Campbelltown Council (including Ingleburn and Macquarie Fields, these suburbs have only minimal connections by both road and public transport to the Sutherland Shire end of the district).
Ah yes I did think about that one. It’s not quite so obvious that this minimises voter movements but I think it probably does.
For Hughes, it’s probably the fact that they tend to just extend Hughes west across the Georges River to take in whatever parts of Liverpool or Campbelltown LGAs don’t fit in other seats and are needed to make up the numbers.
Whereas perhaps a more community-based outcome might have been to push it north into the western half of Banks (East Hills, Milperra, Padstow, etc). But this would then require Banks, Barton, Watson, Fowler, etc to undergo more significant voter movement.
So Hughes is the bits-and-pieces type seat so the others can stay relatively untouched.
Fair point Mark, in this century Hughes has tended to be the ‘leftover’ seat with no clear anchor. It is probably akin to the state seat of Holsworthy/Menai which is also seen as a mixed seat that straddles the Sutherland Shire and Liverpool council areas.
@Ben Raue
I found the new McMahon to be incredibly weird mashing up Blacktown with Fairfield.
It was already weird before with Minchinbury St Clair and Oxley Park in it but that last redistribution made it way worse.
Considering the issues with KFS and Hughes at the same time, I wish they just made McMahon into a full SW sydney seat based on Fairfield, made a new seat in NW Sydney, and chopped an inner/south sydney seat.
In my proposal I abolished Grayndler (on top of NS), and made Barton into a Wolli Creek-based seat.
@redistributed The growth points are the Sunshine Coast, Ipswich and Logan, and the northern part of the Gold Coast (or more correctly the state seat of Coomera). The Gold Coast as a whole is a little bit light but the southern suburbs of Brisbane and some of the city of Redlands are significantly under. That isn’t matched (yet) in the northern suburbs. My thoughts are that you could move Petrie out of Brisbane and entirely into Moreton Bay and drop Blair down further into Ipswich. That means everything north of Esk that runs behind the seats of Dickson, Longman, Fisher and Fairfax gets added in, including Noosa coming out of Wide Bay.
I’d really like to see the Central and North Queensland seats go back to containing entire communities instead of putting bits of multiple regional cities into each of the seats.
I think melbourne crossing the Yarra.
@mark I agree on trying to move Petrie out of Brisbane however on current seat count Oxley would move further into Ipswich meaning Blair will likely need to take part of Lockyer Valley
@Ben Raue
Regardlng Victoria, there were definitely a lot of awkward boundaries created, though I think the majority of these were due to lack of time/care/imagination rather than for the explicit purpose of minimising elector shifts. If anything, they moved more electors than was necessary in a lot of places.
The decision to abolish Higgins was sound though, and because Victoria has had multiple recent redistributions there are no comparatively lazy seats like Hughes and McMahon as others have noted. The main issue is where they’ve unnecessarily cut across LGA boundaries when better alternatives exist or where they’ve just picked a bunch of random SA1s and called it a day.
Areas where they might have taken a shortcut either due to complexity or minimising elector shifts:
– Splitting Campaspe Shire between Bendigo and Nicholls (or the original proposal to split Hepburn Shire), when it would have been possible to return Woodend to Bendigo with only some minor adjustments to the boundary between McEwen and Scullin or Jagajaga/Casey
– Pushing Casey into Nillumbik Shire when its shortfall could have been addressed by taking in Warrandyte or Emerald (although this would have caused challenges with keeping La Trobe in tolerance)
– Pushing Melbourne into Stonnington without first taking in Southbank
Areas where they could have actually moved less electors for an arguably better result:
– Western Wyndham Council into Corio/Hawke instead of using the clearer state district boundary through Tarneit
– Spotswood into Fraser
– Melbourne Airport into Hawke
– Splitting Campbellfield and Bundoora where Moonee Ponds Creek and Plenty Road would have been a good alternative solutions for Calwell and Jagajaga
– Transferring electors in a chain from Holt to Bruce to Isaacs to Dunkley to Flinders (and splitting Cranbourne, Dandenong, Chelsea and Mount Eliza) instead of just transferring Pearcedale and Tooradin into Flinders
Western Australia was similar in that there was a less costly option to leave the rural divisions largely intact while drawing Bullwinkel fully within the Perth metropolitan area that they didn’t notice.
And a fun fact for those interested in the Victorian redistribution…
The 100 or so electors in Keilor North have changed division at every one of the last 5 Victorian redistributions, thus have been in a different division in every single election since 2016. There’s a reasonable chance that this area will again be moved into a new division at the next redistribution if a new seat is created around Sunbury.
– 1996/1998/2001: Calwell
– 2004/2007/2010: Gorton
– 2013/2016: Calwell
– 2019: Fraser
– 2022: Gorton
– 2025: Hawke
Angas, earlier in this post lists the current population quotas for each state and Victoria is still below 38, let alone anywhere near 38.5 so the chance of the state getting its 39th seat back for the next parliament is extremely slim.
@ Yoh An, the next figures will be released on Friday, last figures showed a decline in Victoria, it will be interesting if the next ones show the same result, The figures will represent the population from 6 months ago
Ben
It might be worth looking at combinations of seats as well. In outer eastern Melbourne there were reasonably minor changes to Casey and Aston – they gained electors and in La Trobe only lost electors. The substantial changes were shifted to Menzies and Deakin. Casey could have not gained Wonga Park – more in common with Menzies and Deakin could have kept Bayswater North if Aston had gained electors from Casey in Upwey etc. maintaining community of interest along the Belgrave train line etc. This was a case in minimising disruption in three seats to concentrate it in two others rather than having an even amount across 5 seats with possibly better community of interest outcomes.
Thanks, the new post is up now. I reference five examples, that’s plenty. I’m sure there’s others.
Ben
Another example – and it directly leads to your example of Kingsford Smith – is the reticence of the AEC to abolish more than one electorate. My submission to the AEC suggested abolishing a North Shore seat (as occurred) and another seat – my suggestion was Grayndler – and then creating a new electorate probably in the South Western Suburbs. Abolishing Grayndler would have allowed Kingsford Smith to move toward Tempe and South Sydney, Wentworth to move south, Sydney to go west, and Barton basically straddle the Cooks River. Hughes would not have ended up in Ingleburn but cross the Georges River. Abolishing two seats would have disrupted a lot more electors but not doing it produced sub optimal outcomes.
Based on the ABS data that came out on Friday, the electorate apportionments per state seem to be unchanged. Qld seems to have slipped slightly further away from the magic figure of 30.51
What’s it at now and it only needs to reach 30.5
Google is your friend
https://www.abs.gov.au/statistics/people/population/national-state-and-territory-population/mar-2025#states-and-territories
Using this data, and applying the formula for the entitlement of the states (state populations divided by double the number of state senators):
NSW 46.12
VIC 37.92
QLD 30.36
SA 10.20
WA 16.29
TAS 3.09 –> 5.00
The last time I did the calculation it was in the low 30.4 s. Unlikely to get there unless something drastic happens – like Victoria or WA slowing riggt down.
I did the calculation December numbers released in June it was 30.36. So it hasn’t changed in. 3 months. Qld will likely get it’s 31st seat in 2029 along with Wa should get it’s 17th seat come 2029. Vic maybe the same but might not be til 2032. Nsw would probably lose a seat though. But I imagine there will be an expansion of parliament before that happens
Could WA somehow pull off the 27th seat next year ? Seems to be increasing fast. Might it just get to 0.51 in 3 quarters time?
Sorry I meant 17th seat!