That the Labor government of Queensland will be defeated on 26 October is an event I have been predicting since the middle of 2023. See my Switzer Daily articles of 12 December 2023 “My Australian predictions for 2024” My Australian predictions for 2024 - Switzer Daily and 28 March 2024 “Queensland Libs do well – disaster elsewhere” Queensland Libs do well – disaster elsewhere - Switzer Daily However, I am presenting this article in a slightly different way than was the case for my most recent predictions for the Northern Territory. See my Switzer Daily article of 2 August 2024 titled “Is Labor set to lose power in the Northern Territory?” Is Labor set to lose power in the Northern Territory? - Switzer Daily
In the NT case, the Mackerras Pendulum (see above diagram) I provided was of my own artwork done beautifully by the Brisbane firm Digital Synergy of Navigator Place, Hendra. My explanation is that I do pendulums for nine jurisdictions and offer them to newspapers who may be interested. In the NT case, no newspaper was interested. However, for the Queensland election Brisbane’s “The Courier Mail and Regional Queensland” was interested so my pendulum here exactly reproduces that published in the that paper on Tuesday 1 October on page 6.
The article in question is re-produced exactly below. Before I do that, however, I should note that I do have a fan club, and some members have disputed individual predictions. I note these now. A Labor supporter living in Redlands disputes my prediction that the LNP will gain that seat from Labor. A Labor supporter in Greenslopes disputes my prediction that the Greens will take that seat from Labor. On the other hand, an LNP supporter in Cairns agrees with my prediction that Katter’s Australian Party will win Mulgrave but disputes my prediction that Labor will hold Cairns. He asserts that the LNP will gain Cairns from Labor with the new LNP member being Yolonda Entsch, wife of the local federal MP Warren Entsch. We shall see.
I may well be wrong in some individual seats but for Queensland as a whole I think my numbers will be roughly correct. So, in addition to “cancellation of deviations” I’ll be able to claim that “cancellation of errors” operated. Meanwhile, a prominent Queensland journalist agrees with me that there will likely be 50 seats in the new parliament for the LNP but thinks the fifth seat for Katter’s Australian Party will be Cook rather than Mulgrave. He thinks the LNP will win Mulgrave.
Anyway, the article in “The Courier Mail” had this headline: “Pendulum swings to how LNP will triumph” and there was a photo of me accompanied by this description of the article: “Legendary Australian political analyst Malcolm Mackerras looks at the Queensland poll”. The remainder of this Switzer Daily article is a precise reproduction of that in “The Courier Mail”.
The forthcoming election in Queensland is unusually easy to predict. I say that as an 85-year-old man who has been publicly predicting Australian federal and state elections since 1969 – and I have done so by employing four tools. They are the opinion polls, the betting odds, by-election results and what I call my “laws of electoral history”. All four tell me that the LNP will win the October 26 general election, and it will be a majority government.
The problem lies not in forecasting the above – because virtually every pundit is saying that the LNP will win. My problem lies in my (more-or-less) unique willingness to go into details. There is a wide range for error in such an exercise. Anyway, here are my details. I predict 50 seats for the Liberal National Party and 43 for the combination of everyone else. That combination is made up of 31 for Labor, five Greens (South Brisbane, McConnell, Maiwar, Cooper and Greenslopes), five for Katter’s Australian Party (Hill, Hinchinbrook, Mirani, Mulgrave and Traeger), one for Pauline Hanson’s One Nation Party (Keppel, not Mirani – see below) and Noosa for the independent sitting member Ms. Sandy Bolton. (The member for Mirani is Stephen Andrew. He was elected as ONP but has switched to KAP. That is why I think Mirani will switch from ONP to KAP.)
My 50 seats predicted for the LNP are all 35 shown on the right-hand side of the accompanying pendulum as LNP-held plus these gains from Labor: Bundaberg, Nicklin, Hervey Bay, Caloundra, Barron River, Townsville, Thuringowa, Redlands, Mundingburra, Aspley, Pumicestone, Redcliffe, Cook, Pine Rivers and Gaven. From the data, readers can work out which 31 seats I predict Labor to hold but I should note that Mulgrave is being vacated by the retiring Labor member, Curtis Pitt. I predict KAP to take this seat from Labor.
In the above exercise, my tool has been the diagram known as “the Mackerras Pendulum”. I have drawn some 200 of these in my day covering every federal election since 1972, every New South Wales election since 1973, every Victorian election since 1973 and Queensland, South Australian, Western Australian and Northern Territory elections covering almost every election over the same period.
Two problems have arisen for me. The first is the proposition that there is never a uniform swing – and my pendulum is based on the notion of a uniform swing. I agree that there is never a completely uniform swing, so my answer is to deny that my forecasting is based on such an assertion. On what, then, is it based? The short answer is that it is based on cancellation of deviations. So, for every more marginal seat that is retained for local or candidate reasons there will be an apparently safer seat that will fall – and such deviations will cancel out. In that sense my pendulum works, time and again. It is true that my predictions are sometimes wrong. When that is so it has (almost) always been the result of making a wrong guess at the size of the overall percentages of the vote.
My first problem has been raised since I started this business in the early seventies. My second problem is more recent. It is caused by people who say: “In the seventies, eighties and nineties there was a two-party system. Now the swing to minor parties is so great it wrecks any analysis based on the assumption of a two-party system.” My answer is to re-cast the two-party preferred vote. Whereas last century it was Labor on one side and Liberal Party-National Party on the other, it is now Labor-Greens on one side and LNP, KAP and ONP on the other.
The percentages shown on the left-hand side are those swings needed by the LNP to take the seats from the current party. So, for example, the LNP needs a swing of 2.1 per cent to take Hervey Bay from Labor and needs 18.5 per cent to take South Brisbane from the Greens. The percentages shown on the right-hand side have a similar logic. Thus, Labor needs a swing of 1.3 per cent to take Burleigh from the LNP and 19.3 per cent to take Hinchinbrook from KAP.
Finally, the pendulum reflects the 2020 general election results. I have ignored by-elections, except in one case, Ipswich west. In 2020 it was won by Labor and appeared to be a safe seat. The by-election, however, recorded a swing of 18 per cent to the LNP. So, the two-party preferred votes at the by-election held on 16 March this year were 15,801 for the LNP and 13,732 for Labor. Hence its placement on the pendulum.