What might an expanded parliament map look like?

22

A lot of people would love to know how the electoral map would change if the parliament was expanded.

We’ve already been able to get some idea of what might happen by looking back at the historical experience from the 1948-49 and 1983-84 expansions. But the political system has changed since 1984, as has the map.

So I wanted to try drawing my own map, attempting as best as I can to draw sensible boundaries that fit within population constraints, and then examine what those maps would mean.

I decided to draw a map for a House of 200 seats, which would be achieved by expanding the Senate to 16 senators per state.

I drew the map using two different tools. The Australian Redistribution Tool by JWood (a commenter on this blog), which was inspired by another tool by commenter Angas, is a very easy tool to use, particularly for small jurisdictions. For some of the biggest jurisdictions, I instead chose to use QGIS, with the Statto Redistricter plugin.

Unfortunately I don’t have projection data for most jurisdictions, so I used JWood’s estimates based on the number of votes cast at each SA1 at the 2025 election. I applied the stricter 3.5% allowance. These aren’t exactly the same as the actual numbers that would be used if a parliamentary expansion were to take place in 2026, but it gives us some sense. I also realise that I ended up using the current population numbers, not the projections, for South Australia, as I was used to using those numbers for other states. The whole thing is a thought exercise, but I have applied strict numerical limits regardless.

A lot of caveats are needed. For a start, my map is one of an infinite range of possible maps. I have tried to draw compact and logical maps, but there may be places where others with better local knowledge could draw them better. Where I have created new electorates, I have usually named them after local geography, to understand their location more easily. In a few cases I have resurrected abolished seat names like North Sydney, Higgins and Gwydir. In reality it would be a significant task to identify at least 50 honoured Australians to have new electorates named after them.

The new boundaries are based on combining SA1s, but in some cases a boundary would not follow the SA1 border. To take an example, the national parks to the south of the urbanised Blue Mountains is part of one huge SA1s. So on my map, the southern border of Macquarie runs just to the south of the urbanised part of the Mountains, when in reality it would follow the LGA border.

I should also note that I have not made any changes to Tasmania. We know Tasmania is undergoing a substantial redistribution, but it has nothing to do with the expansion of Parliament.

If the Senate was expanded to 16 senators per state, that would produce the following numbers per state:

  • NSW – 62 (+16)
  • VIC – 51 (+13)
  • QLD – 40 (+10)
  • WA – 22 (+6)
  • SA – 14 (+4)
  • TAS – 5 (-)
  • ACT – 4 (+1)
  • NT – 2 (-)

The ACT barely qualifies for a fourth seat if the parliament is expanded to this size. The Northern Territory’s two electorates, which are currently some of the smallest seats, would be closer to the average, while Tasmania’s over-representation would be reduced.

This map can be toggled to see the differences between the old and new boundaries.

This next map shows the location of newly created electorates. Understandably, they are spread all over the country.

Each state gained roughly one new seat for every three existing seats. While a lot of these seats are in the big cities, there are plenty of new rural seats too. The largest rural seats have tended to get smaller, but some of them are still quite large.

You can also toggle that map to see which seats are marginal (under 6%) on the actual 2025 map, and on my alternative map. Most of the marginal seats are in similar areas but there is some variation.

So with all these caveats in mind, this is my estimate of how the 200 seats would have split based on my boundaries (using 2025 votes).

The Labor government would hold 120 seats, which is 60% of all seats. That is the equivalent of 90 seats in the current House, so they gain slightly less than their proportionate share.

The Coalition increases from 43 to 61, with the Liberal Party benefiting slightly more than the Nationals.

The Greens and Katter’s Australian Party each gain a second seat, and four extra independents are elected.

Overall the ALP increases their numbers by 27.7%, the Coalition by 41.9%, and the crossbench by 46.2%. It may be that the dominant party just doesn’t end up being quite so over-represented in a larger chamber, but I can’t say that with certainty.

This next chart is a replica of one that I made for the 1948 and 1983 redistributions. It compares the shape of the two-party-preferred vote – as the Coalition 2PP increases, what proportion of the House’s seats would be held by the Coalition on a 2PP basis?

The two lines look very similar, although some of Labor’s safest seats are slightly less safe.

Since the 1980s, politics has moved away from the two-party contest. So this next chart also attempts to map out the distribution of two-candidate-preferred margins.

The Coalition accumulates rather a lot of marginal seats, while Labor’s gains are further up the pendulum. Crossbench gains are also in the most marginal seats. This may also explain why Labor’s gains are less than for the other groups.

With the creation of 50 new seats, there is a long list of seats that go to a particular party, and some existing seats flip from one party to another.

  • Labor picks up Belconnen, Bankstown, Blacktown, Epping, Glenfield, Illawarra, Lake Macquarie, Pendle Hill, South Sydney, Terrigal, Brisbane East, Deception Bay, Ipswich, McDowall, Yeerongpilly, Mitcham, Prospect, Tea Tree Gully, Belgrave, Caulfield, Dandenong, Fawkner, Frankston, Nillumbik, South Barwon, Sunbury, Wyndham, Cockburn, Kalamunda, Kwinana, Stirling and Wanneroo
  • The Liberal Party picks up Camden, Hawkesbury, Gawler, Cardinia, Higgins, Murrindindi and South-West
  • The Nationals pick up Coffs Harbour, Goulburn, Great Lakes, Gwydir and Loddon
  • The LNP picks up Buderim, Burnett, Lockyer and Southport
  • Independents pick up North Sydney
  • Katter’s Australian Party pick up Mount Isa

Seven seats flip between parties:

  • Independents pick up Bean and Fremantle from Labor, and Lyne from the Nationals (although the winner is the former Cowper candidate)
  • The Greens pick up Wills from Labor
  • The Coalition picks up Hughes, Forde and Bendigo from Labor

While some MPs may see their seat flip under their feet, there is usually another nearby seat for them to jump into. The Labor MPs for Wills, Hughes, Forde and Fremantle would have the option of moving to Fawkner, Glenfield, Brisbane East and Cockburn, and the Nationals MP for Lyne could move to Great Lakes. There would be no nearby Labor seat near Bendigo, and it’s unlikely the Labor member for Bean would jump to the new seat of Belconnen. But overall, MPs would have a lot more options, with many having a choice of moving to a new seat.

To give some examples:

  • The seat of Banks becomes substantially safer, but the local MP could also jump to the much safer seat of Bankstown, which takes in a big part of Banks.
  • The Labor MPs for Menzies, Deakin and Aston would have the option of moving to Nillumbik or Belgrave.
  • The Liberal member for La Trobe could switch to the slightly safer seat of Cardinia.
  • The Labor member for Sturt could switch to Tea Tree Gully.
  • The Labor member for Moore could switch to Stirling or Wanneroo.

Finally, you can see the partisan distributions on this next map, and you can toggle between the 150-seat and 200-seat map.

The parliamentary expansion doesn’t radically reshape the map. The Liberal Party’s problems in urban Australia don’t go away. They do pick up the seats of Higgins and Hughes, and a new seat in each of the Gold Coast and Sunshine Coast, but mostly Labor and crossbench domination of the big cities remains.

Once again I should emphasise that all of this is hypothetical – the precise population numbers won’t be what I have used, and the mapmakers will likely make many different decisions. But I think this exercise gives a sense of what may happen if parliament was expanded.

Liked it? Take a second to support the Tally Room on Patreon!
Become a patron at Patreon!

22 COMMENTS

  1. Fascinating work Ben. Two points:
    1. The seat of Ipswich does not actually contain Ipswich – Springfield would be a more appropriate name.
    2. Gwydir and Murrindindi would be the best examples of “bits and pieces” seats here (along with Lyons)

  2. My issue with the map is the continued splitting of Queensland’s regional cities that fracture any community of interest. The only LGA that needs to be split due to population is Townsville.

    For example, the seat of Dawson contains part of Mackay, most of Whitsunday, part of Burdekin and finishes in the Townsville suburbs. Instead it could contain all of Mackay and all of Whitsunday – including Collinsville which is currently in Capricornia.

    It is a source of endless frustration that these issues are repeated each election while there are endless debates about individual streets in South-East Queensland.

  3. The Victorian suggestions make a lot of sense, great work by Ben

    There will always be a bits and pieces seat and Murrindindi would always be a prime candidate for such a seat as I believe McEwan has been one in the past

  4. Great work Ben!
    For the most part these boundaries make sense, there are some slight changes I would want to make around Deakin the end part looks really odd, Aston I would want the Creek as the boundary and I think having Lake Corangamite in Corangamite would be ideal as well.

  5. Under this map, I can potentially see the Liberals being able to pick up the new seats of Deception Bay and Epping if they run suitable candidates. For Epping, Labor could hold it if Jerome Laxale moves to contest there instead of Bennelong.

  6. This was a lot of fun to peruse! Great work, Ben. This makes a compelling case for an expansion of parliament. Flicking back to the 150-seat map from the 200-seat map… goodness… it really makes you realise just how monstrously oversized the current electorates are.

  7. What an absolute waste of space. Firstly Australia is over represented in Federal parliament and having 16 state senators would result in a totally unworkable chamber. You have also not considered community of interest in your redistribution. Not following LGA boundaries is a prime example. This country needs more States and less representatives

  8. Mark, I’d be interested in your thoughts in how you’d draw those north Queensland seats. I found both on the central and north coast of QLD and the north coast of NSW it is very hard to avoid splitting the big cities (Port and Coffs have the same problem).

  9. @Ben Raue Challenge Accepted Ben! 🙂 The problems starts with Noosa (which is really a Sunshine Coast seat) and then creates dislocations all the way up the coast. Which is why Kennedy wanders around like a drunk uncle at a wedding…

  10. Mark, you can use this site to draw your own map. It works for QLD. https://auredistribution.com/fed_qld_2026.html

    Nino, there’s always going to be LGAs that need to be split but I’ve used a lot of LGA borders! If you have specific criticisms I’d be interested in them, like I said I’m sure a proper process could produce better maps.

    I don’t see how a bigger Senate would be “unworkable”. It’s a bit strange to say we’re over-represented when our parliament is so much smaller than other similar countries.

  11. I do have a few issues with some of the new Brisbane and would like to a see a 175 and a 162 seat proposal. But otherwise great work

  12. Hi! I only had one question with this map, how did you manage to put the vote shares of the candidates on the map as AURedistributions currently does not have that feature, also, how did you merge the maps?

    Any help would be greatly appreciated as I’ve tried doing something like this in the past but I wasn’t sure how, these may seem like slightly stupid questions so I’m sorry about that.

  13. So I decided to go back to first principles and drew a fairly coherent Mackay-based Dawson. I figure if you could make that work with areas further north and west then you could make the other central Queensland seats work. It covers the Whitsunday LGA and the eastern half of Isaac LGA.

    But the areas to the north and west – the parts previously in Dawson, plus Herbert, Kennedy, Leichhardt, and I even added in the western half of Isaac and added a bunch of south-western LGAs to Kennedy, have pretty much 4.5 quotas of population. And Kennedy can’t keep creeping deeper into Maranoa without being absolutely enormous.

    So I think unifying the central Queensland cities would require some pretty crazy map-drawing further inland.

  14. Hi Owen, not stupid questions at all.

    I use R, specifically the tidyverse. The part that merges the SA1s into a single polygon is this ugly bit of code in a for loop:

    for (i in 1:nrow(newelectorates_list)) {
    newelectorate_name = newelectorates_list$ProposedDivision[i]
    temp_polygons = sa1_shapefiles_newelectorates %>%
    filter(ProposedDivision == newelectorate_name) %>%
    group_by(ProposedDivision) %>%
    summarise() %>%
    rename(Name = ProposedDivision) %>%
    select(Name,geometry)
    if(i == 1) {
    new_house_polygons <- temp_polygons } else { new_house_polygons <- rbind(new_house_polygons,temp_polygons) } } As far as the vote estimates - I have calculated the number of votes per SA1 based on the number of votes cast at each SA1 from each polling place. I don't share this data but it is possible to calculate this.

  15. Owen, I was surprised to see the breakup of Bendigo into two makes it worse for Labor. But checking federal polling place results, it makes sense as the best areas for Labor were in the south, around Kyneton and Castlemaine which are combined with ultra safe National voting areas further west. Labor’s vote in Bendigo and surrounds was much lower.

  16. @Yoh An, indeed, it did feel a bit like it split the COI of the leftwing vote in the Kyneton-Daylesford-Castlemaine area, not a gerrymander, because that implies it was on purpose but certainly a bit of a COI split. Particularly because that triangle is so politically different to the other areas in the new Loddon electorate (which I feel should be named Quick because John Quick was basically the only big whig federation person from regional Victoria and he was from the Bendigo area, but that’s just a nitpicky thing).

  17. Agree Owen, this is one of the few places where the community of interest (COI) is split compared to the old map. At a state level, it is Bendigo itself which is split with one half combined with the Castlemaine area as Bendigo West, and the other half combined with Eaglehawk and adjacent smaller communities as Bendigo East.

  18. I will need to study the maps when I can find a hole to climb into with plenty of kombucha and not come out for three days… All of this information, I feel like a mosquito that flew into a nudist camp, I don’t know where to begin…

    That said, aside from the basic argument of size/population constraints, I think the most significant effect of this will be that it will almost certainly force the Liberals into a centrist position. A necessary one, I would say, but it would drive the far right bonkers.

    I can’t wait until Mark Yore is bored again and blesses us with an extended take on where this is all going…

    In general, I am starting to shift a bit away from treating geographical bits as hard borders for communities of interest and toward the commonality of living, driving, work, etc. These days, a bridge means nothing in an urban environment.

    There is a bloke on YouTube who looked at an expanded parliament from the perspective of a uni student voting. He surmised that there are up to twenty seats where students would make up a significant share of the vote.

LEAVE A REPLY

Please enter your comment!
Please enter your name here