South Australian election guide launched


I’ve now completed my guide to the South Australian state election, due on March 19 next year.

The guide features profiles of all 47 House of Assembly electorates, as well as the Legislative Council contest.

For now, I’ve released two of the profiles as a sample. I will release the full guide for free access early in 2022. Those two seats are Elder and Newland.

You can get access to the entire election guide (along with the federal election guide), you can sign up as a Patreon donor for $5 or more per month. Thank you to everyone who is already giving their support – this wouldn’t be possible without your help.

I will be returning to the topic of the South Australian election as we get closer to the election. There are barely any candidates listed on the guide so far – please feel free to post candidate names as they turn up, and I will add them over time. But this is it for now.

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


  1. Correct me if I’m wrong and not that it matters but I believe South Australia is the only state or territory to never have had a woman premier or chief minister.

    If Marshall loses and at this state the election looks 50/50 then Vicki Chapman could take on the leadership, Not to mention about the past Isobel Redmond probably would have won in 2014 had she stayed as leader.

    I still find it strange that Labor has only won the TPP once since 1985 which would be 2006, does anybody know the historical reason why this has happened? Yes Labor was in government in many of those years but you have to remember this would be the worst track record for any state or territorial Labor electoral performances based on the TPP.

    Labor has not won the TPP twice in a row since the 80s either, South Australia had a strange anomaly where the winning party loses the TPP why is that? During the Playford days it was obvious because the boundaries were illegally drawn in his parties favour but considering they are drawn by the independent AEC/SAEC why have the boundaries benefited Labor which is evident because Labor has won government 4 times since 1989 yet lost the TPP.

  2. Two of those elections in the 2002-18 period were Labor minority governments, supported by independents from what would otherwise have been Lib seats (to the confusion of ECSA and the fury of the Libs). Sorry Libs, but if you lose your former premier’s seat to an indie who then supports a Labor govt (and is still there!), that’s not the fault of the boundaries.

    2010 was definitely the fault of a few Labor MPs in marginal seats. How dare they win their own seats on their own personal vote, while the rest of the state goes the other way. Haven’t these people ever heard of a uniform swing?

  3. This topic has come up a bunch of times on this blog and on the podcast. Look up stuff about the “fairness clause”.

    Basically the distribution of votes in South Australia advantages Labor. The Liberals have more super-safe seats that waste votes. Single-member electorates aren’t meant to produce a fair result, so don’t complain when they don’t.

  4. I outlined my thoughts in my 2016 redistribution submission…..but I think the issue in SA is that it’s been a consistent ‘bias’ (if you want to call it that ) towards Labor for decades. If it just a couple of one-offs or specific freak instances, or if it was very close 2PPs, you could say fair enough that’s footy.

    But a party consistently winning 52-53% of the TPP vote shouldn’t be at the whims of Independents or individual swings again and again. That’s normally a very solid win anywhere else.

    I don’t say it’s conscious or deliberate gerrymandering or anything….but I think it was fair to say before 2016 that some bias had slowly crept into the boundaries over time. The 2016 redistribution seemed to do a better job of addressing it than previously.

  5. SA having such a small parliament might be part of the issue. With only 47 seats, 24 for a majority, the break-even zone around 50% 2pp is larger than for other states.

    Re-running the last eight elections and counting the independent seats for the party that “should” have won them (most as you’d expect, Fisher ALP in 2006):

    2018: Lib 51.9%, 27-20 (57.4%)
    2014: Lib 53.0%, 24-23 (51.1%)
    2010: Lib 51.6%, 21-26 (44.7%)
    2006: ALP 56.8%, 30-17 (63.8%)
    2002: Lib 50.9%, 24-23 (51.1%)
    1997: Lib 51.5%, 26-21 (55.3%)
    1993: Lib 60.9%, 37-10 (78.7%)
    1989: Lib 52.0%, 23-24 (48.9%)


    2018: Lib 5.5%
    2014: ALP 1.9%
    2010: ALP 6.9%
    2006: ALP 7.0%
    2002: Lib 0.2%
    1997: Lib 3.8%
    1993: Lib 17.8%
    1989: ALP 3.1%

    Average bias: 1% to the Liberals! Pull out the 1993 outlier, and it’s 1.3% to Labor. 1/47 is 2.1%, so that’s an average bias to Labor of two-thirds of a seat. Removing that might have flipped it the Libs’ way in 1989 and 2014 (if they’d also won Fisher), but it didn’t change any other results.

    The one that sticks out is obviously the 1993 landslide – that’ll happen with single-member seats in lopsided elections. 2006 had just as much bias to Labor as 2010 did… I don’t remember the Libs whinging about rigged boundaries after the Rann-slide. On this measure, 2018 was almost as unfair for Labor as 2010 was for the Libs.

  6. I don’t think the way you’re measuring bias holds up. The single-member voting system will always produce disproportionate results in favour of the winner – the test is whether the median electorate on the 2PP is to the left or right of the state 2PP, which is harder to measure.

  7. Short of some for of proportional representation, which I presume is not on the table for the SA House of Assembly (unfortunately), adding an extra district (for an even number) and also one or more* at large seats to provide a statewide 2PP majority with a more effective means of causing a majority on the House of Assembly. This would however likely require a referendum due to the equal district sizes rule being referendum entrenched.

    *With a majoritarian electoral system, even if multi-member.

  8. Will all respect Tom, I think that would be a very bad way to go.

    2PP is an artificial constructed concept, particularly when you go outside one seat. If the Libs win a majority of the 2PP that doesn’t mean a majority voted for them. I don’t see why they should be entitled to a majority in the parliament. It’s like having your cake and eating it too. Fairness when you only look at Liberal vs Labor but no fairness for anyone else in the voting system.

  9. the 2pp measures seats …. and gives a good indication of how the state support fares….. seems electoral geography means labor can only win 2 at best non urban seats. On the other hand they are completive in many Adelaide seats…. even the premier.’s seat in the right circumstances. ignoring the 2 landslides the margin is in the range 52/48 either may seems to apply. The Fairness clause now gone did not seem to work as intended. I suggest a better way is the existing preferential voting…. followed by a list system to balance the 2pp using pr. Maybe more seats are needed?

  10. The “fairness clause” didn’t work particularly well and it’s especially broken in a context where a third party has prospects of cracking 20%. If South Australians want seat totals to match vote totals, they should adopt a proportional system. It’s that simple.


    A majoritarian multi-seat statewide electorate would certainly have its problems, mostly accentuating victories where both the majority of districts and the 2PP agree. The more seats, the more problem.

    A statewide electorate (combined with an even number of local electorates) would create incentive to campaign in safe seats, as they would also be in a marginal seat. Not as fair as ordinary proportional representation (which is unlikely to be adopted in the House of Assembly, except maybe as port of 2PP MMP) but fairer than the current system.

  12. How popular is Steve Marshall?, is he a Mark McGowan popular premier? I have read the Liberals are having trouble here, why doesn’t he seem to be cutting through? I didn’t think he had done anything controversial.

  13. An even number of seats is a particularly terrible idea. Look what happened in the last Lib govt in Vic: 45-43, with no crossbench, then the 45th Lib left the party and made parliament almost unworkable.


    An even number of local single member electorates with an additional statewide seat that would effectively be a tiebreaker.

    A rouge government MP is just as much an issue in an odd-numbered chamber of government where the government has a single seat majority.

    Victoria in 2010 very nearly had 44-44, which would have triggered a fight over who to appoint as speaker, potentially of a protracted nature. It may have resulted in a referendum to increase the size of the Victorian Parliament (which got referendum entrenched by Bracks`s reforms) by one region (adding 5 MLC and 11 MLAs/LA districts).

  15. It would not be an effective tie-breaker. Consider this: 48 normal seats in parliament, plus an at-large tie-breaker. The Libs win 50.3% of the 2pp vote (and therefore the at-large seat), and 23 of the 48 seats. Overall that’s 24/49, so they win the 2pp vote, the at-large seat you designed to ensure whoever wins the 2pp vote becomes govt… and they’re still the opposition.

    It’s a Rube Goldberg solution to the problem. Over-complicated, and not guaranteed to work.

  16. Steven Marshall is bland pollie with no personality or character and lied that he would not privatise anything, come on guys wake up! Liberals will defiantly not win, labor has a good campaign machine on SATURDAY especially with peter being the leader he is a good communicator and will be a force to beat.

Comments are closed.