After every election, there is a process to determine how many electorates each state and territory is entitled to at the following federal election. That process will take place in the middle of 2023. The process will utilise the latest population estimates from the Australian Bureau of Statistics, but at the moment it appears likely that the total number of seats will drop back to 150, with the two largest states each losing a seat and Western Australia gaining one.
The ABS releases population estimates every three months. The latest figures (as of June 2022) were released at the end of last year. I expect the September and December 2022 estimates to be released prior to the determination of entitlement in July/August 2023.
First up, let’s check out these estimates for each state and territory:
Victoria easily loses its 39th seat which it gained for the first time in 2020, and Western Australia regains its 16th seat which it first gained in 2014 and then lost in 2020.
New South Wales just narrowly loses its 47th seat. New South Wales had been steadily losing electorates since the early 1990s, but hasn't lost a seat since 2014.
It's also worth noting that the Northern Territory's 1.43 quotas is sufficient to retain their two seats thanks to the 'harmonic mean' rule implemented during the last term. Two Northern Territory electorates would be closer to the national quota than a single electorate, so they get to keep the two. They would need to drop to 1.33 to be reduced to one seat. But that rule only applies to territories with less than three quotas.
New South Wales is very close to the line and it's not hard to imagine the state climbing back over the line in the next two population estimates.
To get a sense of trends, I have charted out the last ten population estimates covering two and a half years. For this chart, I subtracted the 'safe' quotas for each state to show how close they are to losing or gaining a seat with what remains of their population.
You can see a break in the trend in late 2021. I won't pretend to fully understand the process of estimating population numbers, but I know that these estimates are based on the most recent census. As we came closer to the 2021 census, the estimates were relying on a census that was almost five years out of date. Once the new census had been completed, there was a jump in the numbers. Those shifts are less than 0.5% of the state's total population, but that can be enough to make a difference at the margins when it comes to being entitlement to one last seat.
Prior to that jump, no state was on track to change its entitlement. After the jump, two states have definitely swapped a seat, with New South Wales hovering on the edge.
There is an added complication in the case of New South Wales. If a state goes seven years without a redistribution being otherwise triggered (such as, for example, a change in seat entitlement), one is triggered anyway, and that seven years is up for New South Wales in February. So New South Wales will commence a redistribution based on its current 47-seat entitlement in February. If New South Wales loses its 47th seat when the entitlement is determined later this year, that first redistribution process will be cancelled and restarted with a new quota.
The chart above also suggests some potential changes in 2026 - Queensland is creeping closer to a 31st seat, while Victoria's 38th seat is also in serious danger. But a lot could happen before then.
I'll be back with another blog post tomorrow analysing the intra-state enrolment quotas within New South Wales, Victoria and Western Australia.
I am wondering – aren’t there a fixed number of seats (144), that are added to (2 extra in Tasmania as it must have at least 5 seats) and 3 ACT and 2 NT seats? So how could the overall parliamentary size be decreasing by 1?
Marko, the seat numbers are not fixed. The 144 is just a guide as I believe the Constitution says something like “House numbers must be approximately double the size of the Senate”. There is no guarantee that all six states (excluding Tasmania’s automatic 5) will match that figure due to population variations. The only way to fix the numbers is to use an iterative method similar to that used for US House apportionments.
As a general comment to Ben, another option is to simply defer the scheduled redistribution for NSW. I believe the electoral laws allow a scheduled 7-year rule redistribution to be postponed if it is likely a state/territory will have its entitlement changed. I think this occurred with the 2014 apportionment, as the ACT was due for a scheduled 7-year rule redistribution a few months beforehand. At the time, it was possible the territory could have gained a 3rd seat, so the redistribution was postponed until the entitlement was made.
The Parliamentary Library publishes an outlook for this based on the latest released figures and a projection for the couple of quarters ahead.
Their latest (which I agree with) shows Queensland missing the threshold by a fraction as our covid-era growth relative to the rest of the country slows back down. We’d also be due for a redistribution shortly before the next election anyway, so that will be postponed both for the election and then for an additional year with the entitlement determination, resulting in a fairly major redistribution in 2026-2027.
There is a strong argument for more seats. If the increase Waa justified in 1984 it is surely justified now. 100000 plus is too big
Just to explain the thing about the number of seats – yes the target is 144 seats for the six states but it is affected by rounding. So if you put aside the 5 territory seats and Tasmania’s two bonus seats, we currently have exactly 144 state seats by the typical allocation method. But if you look at the table above, four states round down while only two round up.
Under the current rules we’ve had as few as 147 House seats in 1993, and 148 at quite a few elections. Back in 1993 there were 3 territory seats. Plus Tasmania back then had 3.95 quotas. So that would leave 143 typical seats, suggesting that, like in the current situation, 4 states rounded down and 2 rounded up.
Over time the House has slightly expanded but the causes of that appear to be:
-The Territories gaining more seats – was 3 in 1984, jumped to 4 briefly in 1996, 3 in 1998, 4 in 2001 and 5 since 2019.
-Tasmania falling further behind the mainland states. Used to be 1 bonus seat, but is now 2 bonus seats.
-Short-term fluctuations due to rounding – typically settles on 144 but could go to 143 or 145.
Thanks for looking at these Ben, I hijacked the Coomera 2020 thread last week going off on a tangent about these estimates. So I appreciate your timely contribution to this topic.
My crude quota estimate arrived at a bit closer to 30.4 – 30.5, but I decided to be a bit more bullish than ABS in my projections. So I still believe that there is an outside chance Queensland ends up with its additional seat.
To provide context, in the Coomera thread we were also speculating on the likelihood of a Queensland state redistribution prior to the 2024 State Election due to rapid population growth in divisions like Coomera. Coomera is currently 34.43% beyond the average divisional enrolment. Updated enrolment figures for January are expected to be released soon.
I have a solution, why not expand parliament to 200 or 250 seats? Australia is fast growing and electorates on average are starting to become overcrowded and it becomes impossible to contact your local MP (much like the USA) although we aren’t anywhere near close to the average population per district that they have.
The UK has 650 seats with 65 million people or so. We have 25 million or so. It’s time to expand parliament. When parliament was last expanded it was under 20 million. Eventually you are going to end up with 250k+ people per seat and that is just way too much responsibility on an MP’s shoulders.
It is also time to expand the senate and give NT and ACT the 12 senators each that they deserve.
Thanks Ben.
Expand yes proportionate to what happened in 1984. To give nt 12 senators does not make sense. This would create a bigger imbalance. There is a nexus between the hor and senate approx 2 to 1 so might get 15 seats each state and 3 each act and nt
Mick but why should Tasmania get 12 senators with a similar population to ACT with 2? Territories shouldn’t get “less rights or representation” just because they have a different status. They are Australians like everyone else.
Daniel, I believe the territories already have ‘less’ rights than the original six states because legislation passed in either ACT or NT can effectively be ‘overruled’ or vetoed by the Federal Parliament. It was only recently that Labor with support of Greens and some Coalition senators passed a bill that enabled the territories to create their own laws for voluntary euthanasia, which was banned previously during the Howard government.
While preparing for the scheduled NSW redistribution, I did my own sums on this a few weeks ago. I had to go back and check to make sure I did the sums right.
So both WA and VIC are teetering on the line, with determinations gaining and losing seats back and forth. WA have gone from 15 to 16 back to 15 and now back up 16 again since 2014, while Victoria goes from 37 in 2017 to 38, up to 39 and now back down to 38.
I didn’t see NSW losing out again, It’s gone from 50 at it’s peak losing Gwyder, Reid and Hunter (with Reid and Hunter renaming Lowe and Charlton respectively). Now they’re probably set to lose another one.
Seeing how close NSW is to the line though, I am certain they will use the provision in the Act to defer the scheduled redistribution until after the determination is made.
That will then mean it will need to start within 30 days of the determination and those of us that like to complete full statewide redistribution submissions will have our work cut out for us doing three states, including the two largest states, at once.
If we adopted a “Tasmania Rule” today (i.e. legislated an increase in the size of Parliament such that Tasmania was entitled to 5 MHRs on a population basis) then I think that would imply 18 Senators per state (and with it a clean increase to 3 Senators per Territory).
AlexJ, I’m in favour of a Tasmania Rule. We could go to 18 senators per state, or keep it at 12 and add a couple of new states. North Queensland and the Riverina would both be good options.
@Wilson, any new state is not guaranteed to have the same number of Senators. Only the Original States get 12 Senators each. Any new state sold almost certainly be allocated an amount of Senators the same, but it’s not guaranteed. Then again, neither ACT or NT where given the same allocation, so maybe there is precedent.
The December numbers have pushed Queensland closer and nsw further from their next seats. Vic and nsw are sure to lose a seat and wa to gain one. Qld is and outside chance to gain one. I’m tipping Blaxland in the western suburbs but if it’s in western Sydney Labor have got problems come next election. In vic I’ reckon it should be macnamara but wouldn’t be surprised at Gellibrand. The wa seat will be in Perth without a doubt but havent had time to draw where. Given that wa was won on the back of ,ark McGowan. Labor have problems holding their gains. And if they lose those they lose their majority.
anyone willing to support a call to rename Werriwa to Bradman? Werriwa no long emcompasses the lake it was name after and therefore dont believe it has meaning now. i think australias great sporting hero deserves an electorate named after him
I don’t see them renaming a seat like Werriwa. It’s a federation name. We’ve also got heaps of seats named after white blokes and the AEC clearly knows it. I don’t think a name like Bradman (or Molan) has any chance.
Ben Raue, if famous non-politicians like Bradman want to be honored then their names could easily be used to replace geographical named state districts. I think SA and Queensland use many non-geographical names for state districts to honour many people who were not as well-known nationwide.
@yoh werriwa is a geographical location that no longer is part of the electorate
True Ben, but the AEC criteria has a policy of retaining original federation names where possible. Granted, they have moved away from this policy with the renaming of Melbourne Ports but that is a more straightforward case of removing a non-indigenous location name. The AEC still retained the names of Werriwa and Corangamite because they are indigenous location names.
Yoh An, the AEC has removed a Federation name in every state except Queensland in recent years.
Melbourne Pirts, Wakefield, Dennison, Kalgoorlie and Gwyder werr all Federation names.
The last two Vic redistributions, the proposed changes tried to remove Corangamite, first to Cox then to Tucker.
Both times there was a very vocal minority that convinced them to keep it, only partially because it was a Federation name. By far the most common reason argued for keeping Corangamite was that it was the status quo.
Werriwa is different because the name in english doesn’t match. Most people would be none the wiser that Werriwa means Lake George.
Still, Federation names, and geographic Federation names in particular shouldn’t be retained when they’re no longer appropriate, purely on the romantic notion that men 120 years ago thought it was a good name. NSW in particular will run into issues having now 6 PM (7 counting Cook) to find a division to name after.
Probably agree Darren, I forgot about the other examples. In particular Denison and Wakefield were name changes made in the most recent redistributions for their respective states.
Although I read on AEC website that federation names for SA and Tasmania are defined in a unique fashion because those states elected their MP’s at large for the first 1901 election so technically those states did not have true ‘original’ federation district names. Thus, it gives the AEC more leeway in renaming.
Personally, I think the whole ‘retain federation district names’ is a weak criterion and should be removed, especially as it conflicts with some of the other criteria such as ‘avoid giving federal seats the same name as a state district’.
@ Yoh An and Darren. for starters it was proposed 7 years ago that Werriwa and Macarthur should be renamed Whitlam and Bradman. Whitlam ultimately took over a different seat. secondly Corangamite had two different name change proposals. Cox and Tucker. Cox was not taken because its obviously sounds like a male apendage and Tucker was disregarded as it could easily be vandilised into a certain swear word.
I would say Denison and Wakefield were retired since they were colonial officials. Furthermore they are technically not Federation seats rather date from 1903 as during the first election SA and TAS were elected at-large.
for those who dont know weereewa as its spelt is over 200km away from the electorate of he same name.
thats just like putting the seat of melbourne in north east victoria where indi is
gippsland i should have said
since new numbers will be released in june wont they determine them on these? by that time qld will easily have an extra seat
when qld gets its next seat they should name it Irwin
According to Cube Root Law, Australia should have around 300 seats in parliament, in other words, this country is severely underrepresented in parliament. 170,000 people per seat is bad for democracy. We should be looking to Canada as an example because they are almost spot on with the Cube Root Law.
There was a post on FiveThirtyEight that was made a couple of years ago that got bumped and put at the front of the website again, it explains more information about the law.
Parliament should pass a law and the AEC should act on it. Why don’t the politicians want more seats if it means more colleagues of their own anyway?
First they’ll need to build a new parliament. No way you’ll fit 300 mps in that chamber
APH does have room to grow, maybe not to 300. As a starting point it would make sense to expand to 14 senators per state which would add about 25 more House seats.
Is there a law there the senate has to be half the seats as the HOR? It’s rather strange and should be changed. Upper chambers are supposed to be smaller (like the U.S and Canadian senate) although the UK house of lords is larger but that is unelected.
Is the Australian parliament going to be stuck at 150-151 like the U.S house is stuck at 435? I agree perhaps a small increase as we go would be better however there is allot of work to do.
A new parliament won’t be needed, they would simply need to put new seating down. Like the Tasmanian parliament is doing now that they increased their parliament to 35 (they removed seats back in the 90’s)
@Daniel T Yes there is a law. It’s in Section 24 of the Constitution:
SECT 24.
The House of Representatives shall be composed of members directly chosen by the people of the Commonwealth, and the number of such members shall be, as nearly as practicable, twice the number of the senators.
@daniel theoretically if the numbers moved around the country enough we could have 153 at the moments
I have thought we should start with the number of electors per Tasmanian seat (as there needs to be 5), use that as the basis for each seat, then work the Senate from there.
Maybe that is what they do now but don’t know.
Daniel
There was a referendum to remove the Nexus (as it is called) in 1967. It was soundly defeated. The major opposition at the time being from the DLP, some of the Country Party and the smaller states. It has been off the agenda since.
As Ben says an extra 12 senators and 24 MHRs makes sense. And we have had significant population increase.
The states also should look at increasing their numbers of seats. It is particularly ridiculous that NSW now has fewer lower house seats now than it had in 1950.
@Mostly Labor Voter
I think that would cause serious volatility in seat numbers over time. Minute changes in Tasmania’s voting population (relative to Australia’s) can easily cause the parliament to gain or lose many seats in one go, with presumably NSW and Vic taking much of the brunt.
I support an expansion to the parliament, just like Ben and redistributed.
If we could introduce multiple-member seats in HoR while we are at it (Doesn’t have to be huge – 3 members is fine. Also, expanding parliament would be a good timing for this), then I think that’d be great. Major parties probably wouldn’t like this though Labor may like the idea of thwarting the Greens in Richmond (Greens on 25% here, aiming to win on Greens v Nationals 2pp) by chucking in Page (Greens on 8%) territory with it (Assuming 3-member-seats, it will put Greens far from the required 25%)
single member seats is better. no points for second place.
New ABS estimates (Sept 2022 pop) are out. My quota calcs:
NSW 46.43
VIC 37.72
QLD 30.35
SA 10.36
WA 15.90
TAS 3.24
ACT 2.60
TAS 1.42
Main news is that NSW losing a division is now looking very likely. Australia as a whole has been growing at about 0.4% per quarter. If other states maintain that pace, NSW would need to grow at more than 0.6% in the next quarter to keep its 47th seat.
Comments are closed.