WA’s upper house shifting to one big electorate


The current voting system for the Western Australian upper house is set to be overhauled after the WA state government accepted the recommendations of its expert panel earlier today. The panel has recommended the abolition of all regions used to elect the upper house, with 37 members1 to be elected by one statewide electorate, thus abolishing the severe malapportionment that has given far more power to rural voters. The panel also recommended the abolition of group voting tickets, making it easier to vote below the line and allowing voters to mark their own preferences above the line.

The new voting system will be far more proportional and equal, but will lead to a larger ballot and require voters to mark many more preferences if they want their vote to have maximum power.

The report proposes the abolition of group voting tickets, which is definitely a good thing, and will allow voters to number preferences above the line. They do also propose making it easier to vote below the line by not requiring voters to number every box (although I’ll come back to this later).

The panel’s report is worth a read, as they explore the options they had in terms of splitting up the state. They run through various ways in which dividing the state up into multiple regions limits voter equality (most of which are reasonably easy to overcome), but also explore various models. The biggest barrier to simply redistributing the current regions to be equal in size is that so much of the state’s population is in Perth – close to 75%.

You can divide the state into four regions, with one massive nine-member region covering all of rural Western Australia, or have smaller regions with a small number of MLCs which fail to provide much proportionality (they discuss 3-member regions). But once you are electing MLCs to represent a region stretching from the Kimberley to the south-west, you might as well elect MLCs at large.

So they have instead opted to go for a single state-wide electorate, as seen in New South Wales and South Australia. The difference is that those two states only elect half of their upper house at each election, so the quota is higher. The panel does consider electing 18 members per election for a total of 37, and thus giving MLCs an eight-year term, but doesn’t end up recommending it. I don’t see how it could be politically feasible to introduce eight-year terms in a chamber currently elected for four-year terms.

To be clear, I think a low quota is good. Proportionality is good, and I see no issue with allowing minor parties to win a seat off a relatively low vote. The quota would be about 2.63% of the total vote. Electing an upper house via regions would limit the proportionality, and would mean minor parties as large as the Greens wouldn’t be guaranteed any seats. In contrast, a 37-member electorate will ensure at least two Greens MLCs. They probably won’t have a chance of winning five seats (as has happened on occasion) but it’s hard to see them dropping to one, as happened earlier this year.

So what is the problem with electing so many members on one ballot? Basically, the low quota will attract a large number of groups and candidates, all on the one ballot. Larger parties would be encouraged to run at least twenty candidates, and a wide range of parties will be attracted by the prospect of winning a seat off just 1-2%.

A voter who is marking their preferences above the line could number many boxes and still see their vote exhaust. It will be even more difficult below the line to ensure your vote flows to a range of candidates.

The panel has expressed concern about exhaustion rates below the line, and have proposed that voters should be required to number at least 22 boxes to cast a formal below the line vote, which means that below-the-line voting won’t be much easier than it is now.

There really isn’t any system in Australia where below-the-line voting can be used to modify the party’s candidate list in practice. It requires a lot of voters to vote below the line, since preferences cast above-the-line will follow the party’s list. Lisa Singh’s election to the Senate in 2016 required a very high primary vote for a single candidate and a bit of luck.

When you have to number twenty-two boxes to vote below the line, it just isn’t a feasible option for many voters. So in practice this system will be a closed-list system. If you are, say, the tenth candidate on the Labor ticket, you would need 27% of the primary vote to win, and nothing else matters.

If we’re going to elect 37 members in a single electorate, we should consider simpler systems used in other countries. Many European countries use party list proportional representation (list PR) to elect more than 37 candidates in a single electorate.

I personally like the single transferable vote, either with Hare-Clark or with above-the-line voting. I think it works great if you’re electing up to, say, nine members in an electorate. In those situations, preference flows between different parties can make a significant difference, and in some cases the number of seats up for election means many voters won’t be able to be represented by their first choice.

But when you are electing larger numbers in a single electorate, such as the 37 proposed for the WA Legislative Council, or if you were to elect all 42 NSW MLCs in one election, a simple divisor method of list PR would achieve about the same outcome while making the ballot much simpler. When the quota is around 2.6% you don’t need to use preferences to ensure a fair outcome.

You could use a closed list system, but there are open list options where voters cast a ballot for an individual candidate. Seats are allocated to parties based on their share of the vote, and those seats are then allocated to the individuals with the most votes. It gives real power to voters to decide which candidates within a party win seats, a power they don’t have under any system using “above-the-line” voting around Australia.

Yet it seems like Australians are fixed to the idea that a fair voting system needs to use preferences, even when the voting system reaches absurd levels of complexity.

There have been a number of attempts to introduce list PR for elections using a large district magnitude. When the NSW Legislative Council shifted to direct election in 1978, the Wran Labor government proposed using list PR to elect 15 members per election, but compromised for single transferable vote after Liberal/National opposition.

South Australia’s upper house was first elected using universal suffrage in 1975, and used list PR at the 1975 and 1979 elections.

And the first two elections to the ACT Legislative Assembly in 1989 and 1992 used a modified list method which was mangled in the Senate to include elements of preferences that massively complicated the system, and was eventually replaced by smaller Hare-Clarke districts in 1995.

It appears that list PR would violate the Western Australian state constitution, but I still think it’s worth discussing this alternative. An STV election conducted with 37 seats up for election is far from the ideal way to use that voting system.

1 The actual report seems to recommend a 36-seat electorate, but the Attorney-General has confirmed that the legislation will increase the chamber to 37.

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


  1. Haha it’s pretty farcical that another system attractive to tablecloth ballots is going to be introduced, but at least it’s an improvement.

    Judging by the last election vote tallies we would have seen 23 Labor, 7 Liberals, and a crossbench of 7 with 2 or 3 Greens, one each of Nationals, Legalise Cannabis, Australian Christians, and Liberal Democrats with One Nation, and Shooters fighting for the last spot and to oust the 3rd Green. Still a Mos Eisley cantina upper house but at least those representatives would be there due to intentional decisions of voters.

    The 2017 election might be more reasonable to look at. I think it would have been 15 Labor, 10 Liberal, and a crossbench of 12 with 3 Greens, 3 One Nation, 2 Nationals, plus a Shooter, Christian, Liberal Democrat, and Animal Justice. With 19 required for an absolute majority the government has a lot of options.

  2. How’s this? – Optional preferential voting, above-the-line only. Best of both worlds. Essentially list PR where voters can express preferences if they wish.

    Preferences are a means for voters to express what they really want in greater detail. I wouldn’t want to see that taken away anywhere.

  3. If a week is a long time in politics, then 44 years is an eternity. Yet it was 1977 when Bob Hetherington, MLC, introduced a bill for PR, on the basis on one vote, one value, to the Legislative Council, which, inevitably, was unsuccessful. It is worthy of note that the driving force behind the adoption of this policy by the ALP State Conference, was the ALP’s Electoral Reform Committee, lead, with great energy and enthusiasm, by Graham Hawkes and Gordon Payne.
    I was very pleased to note that both Messrs Hawkes and Payne made 2 submissions to the Committee, and that, after such a long and dedicated campaign, their vision is finally coming to fruition!

  4. Why does WA need such a big upper house compared to lower house – 37 to 59 – wouldn’t it make more sense to have say 25 seats in the LC and increase the size of the lower house so that the total amount of representation is not reduced. 37 seats elected statewide will see a lot of useless hacks representing no one in particular and depending on the details micro parties on low votes will still get elected. It seems that this has not been thought through.
    Good to see that Group Voting tickets will go – Daniel Andrews when will you join in on abolishing group voting tickets?

  5. @redistributed – I doubt Dan Andrews and his government would do it, as the existence of GVTs has meant that the various other minor parties have used it to their advantage to win LC seats at the expense of the Greens, which suits them as the Greens then don’t become the main negotiating path to passing legislation if the Liberals intend to block it. Also, attempting to do so would sour relations with said minor parties.

  6. https://www.tallyroom.com.au/43787/comment-page-1#comment-756115

    The equation for the 2018 election, where the ALP had a reasonable chance of getting a progressive majority with a minimum number of Greens MLCs (which it in fact did) is different from the equation for 2022 when they are likely to loose seats and therefore if want to get progressive legislation through they will need Greens to be elected instead of right of the ALP micros. I think there is a reasonable chance of a last minute switch to ATL preferencing, with the Electoral Matters Committee report saying a separate inquiry from the inquiry into the last election is needed before such a change and the inquiry not being held yet indicating any change would b last minute.

  7. I wouldn’t want to lose preferential voting either, sorry. I research every single independent and micro-party to find the best fit, the least offensive, and give an honest vote as best I can according to my political views. I have quite often given my first preference to candidates who end up with tiny vote tallies, knowing perfectly well they wouldn’t win, but they were the best fit. I wouldn’t want to be put in the position of being forced to change my vote and have to guess who will make quota and who won’t and not be able to vote for who I really want to vote for. Preferential voting is an essential part of democracy.

  8. Lots of functional democracies don’t use preferential voting. I’m not talking about the US, UK, Canada, but New Zealand, mainland Europe, basically none of them use preferences. Preferences are a handy way to improve the effectiveness of votes when the district magnitude is low but when you are electing a larger number it can just add unnecessary complexity.

    I don’t think we should design our voting system around people like yourself. Most people don’t do that level of research and investment of time, and I don’t think it should be necessary to ensure your vote has maximum power.

    And I don’t think the voting system should be designed so that you can vote for parties with no hope of winning, particularly when the level of support needed to win is so low. If a party doesn’t have a chance of polling the 1.5% or so that would give them a good chance of winning the last seat, then I don’t think there’s anything wrong with voters switching their vote to someone more viable.

    You say “I wouldn’t want to be put in the position of being forced to change my vote and have to guess who will make quota and who won’t and not be able to vote for who I really want to vote for”, but in this scenario every party polling over 2% will win a seat. So you only need to consider this point if your vote is for a very small vote. I don’t have a problem with this.

    I would also point out that casting a below-the-line vote has never changed the outcome of a NSW Legislative Council or SA Legislative Council election. The number of valid boxes required to cast an informal vote is simply too many for sufficient people to cast a below-the-line vote for someone to be able to win that way. So while above-the-line preferences may influence the final seat in some cases, below-the-line preferences within a party group won’t.

    On the other hand, a voting system like open list PR would make it easy and viable for people to cast a vote for individual candidates and make a difference.

  9. https://www.tallyroom.com.au/43787?replytocom=756127#comment-756127

    The places in Europe with directly elected presidents use what is effectively a form of preferential voting, the highly inefficient two-round system. France even uses it for its National Assembly (which has single member electorates). Some places in the USA use something along those lines as well. All be it none for large scale PR electorates.

    South Australia has the same minimum 12 preference BTL requirement that the Senate does and allowed Lisa Singh to win with a BTL campaign, with only a marginally larger quota requirement than the DD election where Singh won. The Jim Molan BTL campaign showed that non-Tasmanian BTL campaigns can achieve a not insignificant vote it terms of a single quota. Unless there has been a failed BTL campaign in SA I don`t know about, I would say a lack of trying may well be the potential cause of the lack of BTL victories in the SALC.

    The NSW BTL preference requirement is only 3 more numbers than SA, harder but not necessarily impossible, particularly with a lower quota. Has there ever even been a NSW LC BTL campaign?

    The WA 22 preference minimum BTL may indeed be an insurmountable obstacle, even with a lower quota.

    I am sceptical that a majority of politicians in any non-Hare Clark Australian parliament would be willing to back voters over preselectors with an open list system.

  10. Those European and American examples are all single-member. I agree that preferential voting improves single-member elections, and all the way up to say nine seats in a district I support preferential voting. That’s not relevant to an election with a very high district magnitude.

    Lisa Singh was elected in Tasmania, and she is the only example I have seen of a candidate elected below the line over candidates in their own party. And it required a very high vote in a state used to using Hare-Clark.

    The Senate only requires six preferences, not twelve. Massive difference between that and 22.

    Molan polled 2.9% in a voting system which only requires six preferences below the line. He didn’t get close to winning.

    I agree that an open list system is unlikely, because they have a closed list system now. I’m just saying that is the answer if people want candidate choice, below-the-line voting is a fig-leaf.

  11. https://www.tallyroom.com.au/43787/comment-page-1#comment-756135

    Lisa Singh won in a double dissolution, lost in a half-Senate election. The previous pre-ATL down group wins, under the old pre-ATL system, were in also at a double dissolution (1951). The Molan BTL campaign ran in a half-Senate race where 2.9%+0.2% preferences is way to far from quota to win. There is also not a risk of upsetting a joint ticket with the Nats in SA, unlike with the Molan BTL campaign. Kevin Bonham has said “outside Tasmania, this does show that it is vaguely realistic for a major party candidate to win from outside the ticket order in a double-dissolution” and “if a NSW candidate can poll almost 0.4 DD quotas there is some hope there that someday a win will occur” on the subject:


    (Second last paragraph of the “what does this mean section”.)

    The Senate allows 6-11 preferences as a savings provisions, but instructs 12. The Molan BTL HTVCs had 12 preferences marked on them.

  12. in the 2018 SA election, Jinh Lee who was 4th on the Liberal Ticket didnt explicitly ask for below the line votes but put her own posters up, especially around China Town. She did recieve more BTL votes than other non-lead candidates and did get elected 9th.

  13. There are two ways to achieve regional representation, one is voting for them, the other is through Gerrymandering. These reforms are welcomed and well overdue. Of course the Nats and Shooters Party will be upset because they will not longer win MORE seats despite getting LESS votes than others.

    The previous system was highly undemocratic. Someone’s vote in North West Central across both Legislative Assembly and Legislative Council was worth 9 times as much as mine because of the post code I lived in. Even if you’re sympathetic towards regional vote weighting, you have to admit that previous malapportionment was madness.

  14. The Shooters don’t actually need to be upset – they would’ve won a seat at the last three elections under this system (yes, even the 2021 bulldozer). Turns out trying to get 2.6% across the whole state is easier than 14.3% in particular regions. (Nobody was too upset when they won a seat over a party with a higher vote, because that party was One Nation.)

    I ran the figures for every election back to 2001 under this system: the CDP / Australian Christians would have one seat every single election, the Greens would have a fairly reliable three (four in 2008, only two in 2021), and the rest of the crossbench would’ve been One Nation sparkling and fading, the Democrats (2001), Family First (2005 and 2008), the AJP and LDP (both in 2017), and Legalise Cannabis (2021). Even with the much lower quota, the LDP still owe that seat to the donkey vote in South Metro – flipping 10,000 votes from them to the Libs flips that seat.

    The Nats are the only big losers – never above two seats, and busted back down to one in 2001, 2005 and 2021. Half their problem isn’t big bad city voters trying to rob them of influence, it’s their own voters going elsewhere (One Nation, Shooters etc). If they could get 42% of the vote in the 25% of WA outside Perth, that’s four quotas – it’s not an unreasonable target for them.

Comments are closed.