Federal redistributions will commence this year in three states. Victoria will lose a seat, Western Australia will gain a seat, and New South Wales will also have a redistribution. A redistribution will commence next month for New South Wales with its current 47 seats, but that redistribution may end up being superseded in the middle of the year if New South Wales loses a seat, which is quite possible.
In this post and tomorrow’s post, I’m analysing the intra-state enrolment data to get some sense of how electorates might shift, and which areas might gain or lose a seat, in the upcoming redistributions. I’ve produced maps of all three states, along with tables showing breakdowns of enrolment trends by region.
Today’s post is focused on New South Wales, and specifically explores how such a large gap has emerged between enrolments in Sydney electorates and those in the remainder of the state.
New South Wales hasn’t had a redistribution for seven years, while Victoria and Western Australia were last redrawn prior to the recent election (indeed the change in seat numbers for those two seats simply reverse the changes at the 2020 entitlement). You can see the length of time since the last redistribution in the intra-state statistics – there is much greater intra-state variety in enrolment in New South Wales.
All the maps and tables in this post use the monthly enrolment statistics as of December 2022, but the clickable maps also show you the same statistics as of September, so you can see a bit of a trend.
We don’t know whether the NSW redistribution will be drawing 46 or 47 seats, so we need to analyse both. The 46-seat map will have higher quotas, so each seat will need to include more voters than the 47-seat map.
This chart has divided the state into eight regions, four in Sydney and four outside. I have added up how much the seats in those regions deviate from the 46-seat and 47-seat quota.
| Region | Seats | Deviation (46) | Deviation (47) |
| Hunter & Central Coast | 6 | 12.98 | 26.3 |
| Northern NSW | 4 | 17.22 | 26.3 |
| Southern NSW | 5 | 11.46 | 22.6 |
| Western NSW | 5 | -12.56 | -2.0 |
| Regional NSW | 20 | 29.10 | 73.2 |
| Central Sydney | 7 | -40.33 | -25.9 |
| Northern Sydney | 6 | -50.23 | -38.2 |
| Southern Sydney | 3 | -28.18 | -22.3 |
| Western Sydney | 11 | -10.38 | 13.3 |
| Sydney | 27 | -129.12 | -73.1 |
Overall Sydney is currently over-represented relative to regional NSW. If NSW doesn’t lose any seats, Sydney’s 27 seats only have about 26.27 quotas.
This map can be toggled between showing the 46-seat and 47-seat quotas.
Most of Sydney is currently under quota even if NSW doesn’t lose a seat. The only exceptions are in the outer south-west and north-west.
Macarthur and Werriwa in the south-west are about 27% over quota between them, and are still growing fast. Lindsay, Chifley, Greenway and Mitchell in the north-west are about 23% over quota. That’s half a seat worth of surplus voters in that region.
Across the whole of Sydney, there is a deficit of 73% of a seat, but if you subtract those six outer suburban seats, the rest of the city has a deficit of 1.2 seat quotas.
The most severe deficit is the six seats of northern Sydney, which are 38% of a seat under, with Warringah particularly underpopulated.
No one region is far enough under deficit to abolish a seat just in that area. I suspect a seat around the middle of Sydney (such as Blaxland) could be abolished. Seats in northern Sydney and eastern Sydney will then expand towards the west to fill that gap.
The seats along the regional NSW coast, including the Hunter and Central Coast, are mostly over quota. The Hunter, Central Coast, North Coast and New England are collectively about half a seat over quota, with Hunter and Paterson particularly over quota.
If NSW can maintain its 47 seats, it seems likely that a new seat will need to be created somewhere around the Hunter, pushing seats in Western NSW further south and towards Sydney.
If NSW loses its 47th seat, the higher quota absorbs most of the surplus in regional NSW. The Hunter-Central Coast-North region is only about 0.3 of a seat over now, rather than 0.5. In this case you’d expect to see these seats contract.
The reduction in seats worsens the deficits in established parts of Sydney and reduces the surpluses on the outer edge of Sydney. At this point there is no doubt that a seat in Sydney will have to be lost, probably one in the middle suburban ring.
The sixteen seats in the eastern half of Sydney are collectively 118% under quota, which means you could abolish a seat and absorb its population entirely within the remaining 15 seats and still have too few people to justify those 15 seats.
Indeed across the whole of Sydney the region is 129% under quota, and once you factor in the surplus growth in the six outer suburban seats, the remainder of Sydney is about 1.6 quotas in deficit.
So I think in a 46-seat map we will see a mid-suburban seat such as Blaxland abolished, and despite that we will still see the seats to the west of the abolished seat move further west to absorb the surpluses in outer suburban Sydney, ultimately shifting the south-western Sydney seats further south-west to absorb the surplus in regional NSW.
I was planning now to move on to Victoria and Western Australia but I’ll leave them for another post because I want to answer another question: why is there such a huge divergence between enrolments in Sydney and regional NSW?
My first thought was that this is explain by differential rates of growth since 2016, but that isn’t true.
The final redistribution was based on enrolment data as of December 2014. The election rules require seats to fit within the quota as of the time of the redistribution, but also within a quota based on projected estimates 3.5 years after the conclusion of the redistribution. In this case, projections were created for the population as of August 2019.
On the December 2014 data, the 20 seats in regional NSW had an average enrolment of 106,473, while the 27 in Sydney had 101,264: a gap of 5.1%.
The projections suggested an average enrolment of 110,245 in regional NSW and 110,762 in Sydney in August 2019, a gap the other way of 0.5%. Instead regional NSW had already cracked 110,000 by the time the enrolment statistics began to be reported on the new boundaries in February 2016.
The redistribution had been based on projections of 3.5% in regional NSW and 9.4% in Sydney, but the reality was 10.1% in regional NSW and 8% in Sydney.
So it’s not a story about COVID-19 shifting population away from Sydney – indeed all of these incorrect projections cover a period before the pandemic. This next chart shows how the gap in enrolment shifted over the last seven years.
The gap between regional and Sydney enrolments actually peaked in early 2019, and since then has slightly shrunk, although it is still above the levels in 2016, let alone the projected trends expected at the time.
I'm not sure I fully understand why things went wrong, but the simplest explanation seems to be that Sydney had been growing faster in the years leading up to the redistribution, and the redistribution expected those trends to continue. They did not, and thus the numbers have fallen quite far out of line.
Finally, this chart shows every seat in NSW. Seats are colour-coded according to whether they are in Sydney, and the chart shows how much the seat grew in excess of the projection leading up to August 2019 (negative numbers indicate that actual growth fell short of the projection).
Check the aec website Ben it says it completely replaces it and applies until the next redistribution
How long do the results of a mini-redistribution apply for?
Sub-section 76(6) of the Electoral Act requires that the names and boundaries determined by a mini-redistribution apply until the next redistribution is determined in accordance with sub-section 73 of the Electoral Act or the next mini-redistribution is determined in accordance with sub-section 76(6) of the Electoral Act.
This means that the results of the mini-redistribution will apply until the next redistribution has been triggered and concluded. Redistributions are triggered when:
the number of members of the House of Representatives to which a state or territory is entitled has changed,
the number of electors in more than one-third of the electoral divisions in a state (or one of the electoral divisions in the Australian Capital Territory or the Northern Territory) deviates from the average divisional enrolment by over ten per cent for a period of more than two months, or
a period of seven years has elapsed since the last redistribution process was determined.
https://www.aec.gov.au/electorates/redistributions/mini.htm
Ben, actually Potatoes is correct. The official rules (from AEC website) state the mini redistribution completely replaces the one that was underway, and it remains in place until the next one is triggered (due to malapportionment, 7 years or another change of entitlement).
This link has more details: https://www.aec.gov.au/electorates/redistributions/mini.htm#heading13.
Ideally, because of this rule and the fact that NSW divisions are already out of alignment (several fall outside 10% variance at the moment), the full redistribution should be completed before the next election is held.
@yoh an unfortunately the sae is the case with wa and vic. But if the govt triggers an election there is no choice
Surely the electorates emerging from the mini redistributions would immediately trigger a redistribution due to malapportionment in most circumstances? A Wentworth-Warringah merger would be a monstrosity and if it could potentially happen within the rules, then the rules need to be reformed.
At any rate I don’t see a double dissolution happening. The Greens will likely pass the housing fund well before the 2024 budget (when the fund is due to make its first actual payment).
If I remember correctly (there were some great calculators from mid last year I can’t find any more) if you took the 2022 results and applied it to 12 member electorates, the senate balance shifted right because a lot more PHON and UAP senators get up instead of exhausting. So it’s not in Labor’s interest to do that?
On the other hand Turnbull did his DD high on the theory that he was extremely popular and could secure a dual majority. Albanese may feel the same way and he has more evidence to back it up (e.g. Aston).
John, the malapportionment rule says one third of districts must have more than 10% variance from average enrolment to trigger an automatic redistribution. It doesn’t say anything about the maximum variation of any single district, so you could have one 50% over quota and others all within 10% from quota.
@john the rule says more then 1/3 of districts would need to out of proportion so in this case 16 would need to be in order to trigger the redistribution but given that there are enough that would be or close to being out of proportion atm another redistribution wouldn’t be far off and certainly wouldn’t be 7 years and would most likely be around the time of the next determination at most which is new parliament first sitting date +1 year
And it would be the Warringah-Wentworth merger as the rules state the merged division must take on the names alphabetically:-)
Well that’s a ridiculous rule. I assume if it ever came to pass there would be an outcry to force an immediate redistribution.
Worth noting that in these scenarios the number of seats has changed so you’d expect quite a few seats to be outside 10%, but admittedly in NSW case it doesn’t appear to be based on the numbers I used in this blog post. I count 7/46 seats, and you’d need 16.
Thanks for clearing that up Yoh an and Potatqes. Don’t know where I got the idea a single bad seat could trigger a redistribution from (I think it was with just Pearce being so far off with the WA redistribution that was actually triggered from going 16->15)
So in 46 seat NSW you could theoretically have 31 seats at just over 0.9, 14 seats at close to zero, and the remaining seat at over 17 quotas, and it wouldn’t trigger a redistribution until a single seat slips below 0.9. Not that this would ever happen, just funny to think about.
At any rate the direct link between Warringah and Wentworth seems to be a single Manly Fast Ferry service between Watsons Bay and Manly. There would be a case for water links (even without any ferry services) to count for some isolated communities, but it would make sense for them to only apply if there are no other credible links between those communities and other communities.
Actually john it shifted left as labour maintained is 26 senate spots after gaining 1 in wa meaning they lost 1 elsewhere to the greens and the greens gained 3 total and Jacque lambie gained one in tas as well as pocock in act from the libs giving labour the votes needed to pass the socialist agenda that would not have been passed if the greens and labor stayed the same. They would of needed a combination of Rex Patrick, Jacque lambie or Pauline Hanson If the senate did not shift left.
Cowper needs a redistribution, it’s well over 100,000.
Also guys I’m thinking of starting an election results Fandom wiki. If anyone wants to join it feel free. If I do decide to create it I’ll do it maybe next weekend. Wikipedia blocked me for false accusations of socking and I mostly added election results there so I’m gonna try and do it somewhere else.
I’ll still be active here though.
How strict or sensible will the AEC be if a mini-redistribution must occur?
Looking at pairs of electorates in order of those that are most under-quota, it appears the most under-quota pairings are poor partners.
I can’t see how Wentworth + Warringah could be viable. It would also be very contraversial.
Berowra + Parramatta or Warringah + Bradfield would also be terribly awkward due to the tiny boundary link between what are otherwise very separate electorates.
Berowra + Macquarie looks like a more viable combination, but it’d be a huge area to cover for regions that are so close in to Sydney.
Bradfield + Berowra or Hughes + Banks seem much more viable, even if their combined quotas a slightly less than the other pairings above…
That said, the double dissolution appears more likely to be a threat that Labor is using for bargaining purposes, as it doesn’t seem to be a very realistic proposition. It would risk going to an election during a potential recession (fix federal election terms please!), and it would risk dissolving an otherwise progressive senate for an unknown new mob.
Labor should have more than enough opportunities to do their thing if they negotiate in good faith with the current elected Senate.
@raue there are enough that are close enough that the next redistribution would be before the next election anyway
@john there doesn’t need to even be a link ferry bridge etc. the boundaries only need to touch which they do.
@Ben Raue just on a side note the map of Queensland state electorates is missing Gympie.
@raue it wouldn’t matter if people rioted in the streets over it. The law is quite clear and no amount of crying would force a redistribution.
“@raue it wouldn’t matter if people rioted in the streets over it. The law is quite clear and no amount of crying would force a redistribution.”
An act of parliament would, though. That’s what I’m talking about.
@yea but they have the 1/3 of divisions clause for a reason. By the time people complained and the parliament acted that clause will probably have already been triggered
And besides they can just lay the blame at Albos feet for calling it mid redistribution
If anyone’s interested in seeing my maps or sharing there’s or just general redistribution collaboration
Forgot to mention my email created for this….. redistribution@outlook.com.au
Has there been any data released about how many quotas each state has now?
The shortfalls in Inner and Southern Sydney (up to Blaxland, Hughes and Reid, total of 11 seats) looks like it is currently (May 2023 Gazette) about 10.19 seats (Adjustments made to assume 46 seats in NSW, not 47). I think Reid, Grayndler, Sydney, Wentworth, Kingsford-Smith, Hughes and Cook are quite easy to draw given their proximity to the “corner”.
Grayndler, Sydney, Wentworth, Kingsford-Smith I assume will collectively will extend to the end of Haberfield, Ashfield, Ashbury and down to Cooks River.
The 2 Shire seats I assume will change to take in the whole Revesby chunk and shed most of the St George area (but maybe keep Sans Souci because of School Catchments, specifically Sylvania High). (You cannot get from Moorebank to Menai by public transport without going through Revesby so I think that is a good fix)
Reid may gain the area North of the Train Line in Auburn.
But after these are applied, the remaining seats (around St George area, and Canterbury-Bankstown (minus Revesby) looks like it will be a pain to draw.
@leon Canterbury bankstown can be split tino 2 divisions
Liberals are running pre-selections. Can’t wait for them to pick someone for a seat that gets abolished or re-drawn such that the candidate no longer lives there.
@Dean Ashley
Tbh parachutes arent that bad as long as they aint going from Scotland Island to Cabramatta/Liverpool. So insofar as they are in a neighbouring area it should be fine. That said, it may still have been a good idea to wait a bit more.
As for abolition, my guess is that both Banks and Berowra have to be abolished and a new seat created elsewhere (I personally guess around Marsden Park and the Hawkesbury). Banks because of what I said above, and Berowra because I simply cannot draw a Bradfield without Berowra in it.
The incumbents of those 2 seats are David Coleman who looks like he will leave any time now (esp considering his lengthy leave in the last parliament), and Julian Leeser who looks like can be dealt with if either Alex Hawke gets turfed out or Paul Fletcher retires.
We will get the seat quota confirmed this thursday
Comments are closed.