Back on Tuesday 12 December 2017, an article by me titled “Why John Alexander will win Bennelong” was posted here. In my second article for the New Year, I give my analysis of the New England (2 December) and Bennelong (16 December) by-elections. 

My previous article explained why I was so totally confident in my predictions, notwithstanding that the then current polls for Bennelong were 50:50 between Alexander and Labor’s Kristina Keneally and virtually every other commentator was asserting that Bennelong would go down to the wire.

I also concluded with a promise to do a proper analysis of the Queensland election, held on 25 November. I wrote: “I shall explain why I think this is a pretty good result for the Liberal National Party, contrary to the conventional wisdom of it being otherwise.” 

In this, my first article for the New Year, I do that but precede with some advice for the LNP. I say this to them: the merger of the Liberal and National parties in Queensland was a good idea and it has been a success. Do not go back to the old days by de-merging. I say that, notwithstanding the fact that the LNP has actually lost three of the four state elections since the merger took place.

Consider this piece of history. Peter Beattie in 1998 and Annastacia Palaszczuk in 2015 took Labor from Opposition into minority government, both relying on the independent Peter Wellington who represented a conservative seat on the Sunshine Coast hinterland. At his next election, Peter Beattie took Labor to a landslide win with (on my estimate) 60% of the two-party preferred vote (a pro-Labor swing of 9%) and winning 66 of the then 89 seats compared with 44 at his 1998 win.

By contrast, Palaszczuk has taken Labor from minority government to a narrow majority with no swing at all, largely due to the change in the electoral rules. This result is no triumph for her, just “a win is a win is a win.” Under the rules of Electoral Commission Queensland, there is no official two-party preferred vote, so the analyst must estimate it. 

My estimate for 2015 was 50.8 per cent for Labor and 49.2 per cent for the LNP. For 2017, my numbers are 50.4 and 49.6, respectively. So, any overall swing there might have been was against Labor, contrary to the impressions created by commentary on election night last November.

In 2001, there was no change in the voting rules. There was a redistribution of seats but, on the calculations I did at the time, it made no difference whatsoever to the scales between the two sides of politics. By contrast, this time there was a change in the voting system as well as a redistribution with four more seats in the populous south-east corner of Queensland. 

Late in April 2016, the Legislative Assembly approved a bill to increase its own size from 89 to 93 members and to abandon optional preferential voting while restoring the full preferential vote, the system which has applied for House of Representatives elections since December 1918. To do such a thing was not inherently controversial but it was opposed by the LNP purely because it would benefit Labor. 

At the time, there were people (of whom I was one) who realised that Malcolm Turnbull’s insane “Senate reform” and double dissolution would resurrect the political career of Pauline Hanson, but Turnbull was clearly not among those who understood the madness of what he was doing.

In May 2017, the new boundaries were published, at which point I had no difficulty saying that the notional number of Labor seats on the new boundaries was 48 compared with the actual 44 seats Labor had won on the old maps in 2015. From that point on, there was never any doubt in my mind that Labor would win at least 48 seats, bearing in mind the history of what happens in a case like this.

The sensible way to see this election is to say there are/were two Queenslands, the populous south-east with 61 seats (57 before) and the rest of Queensland with 32, the same number as at the 2009, 2012 and 2015 elections. The term “rest of Queensland” refers to the 94 per cent of its area beginning at Gympie to the north of Brisbane, Nanango to the north-west and the Lockyer valley to the west. 

Due to the change of rules and the map, there is no point in talking about individual seat gains or losses in the south-east. Suffice it to say that the average swing to Labor-Greens in those 61 seats was 0.3 per cent and 45 seats moved in that direction. There were 16 seats recording swings to the LNP or to Pauline Hanson’s One Nation Party where it came second to Labor rather than the LNP.

In terms of seats, the south-east result this time was 36 Labor, 23 LNP and one each for The Greens and the independent in the seat of Noosa. Last time it was 30 for Labor, 26 for the LNP and the independent Peter Wellington making it up to 57. I think readers will agree that such was a pretty excellent set of seat numbers for the left, given that there was only a miserable swing of 0.3 per cent in their direction.

For the rest of Queensland, there is some point in talking about actual seats since there remains in place (more or less) the same boundaries as before. Labor lost two seats, Bundaberg to the LNP and Mirani to the ONP, reducing it from 14 to 12. The LNP remained at 16, gaining Bundaberg but losing Hinchinbrook to Katter’s Australian Party. In terms of seats KAP was the big winner. It held its two former seats but also gained Hinchinbrook from the LNP. Overall the swing against Labor in that part of Queensland was 2.9 per cent.

So why did Labor win? Essentially it was for the reasons advanced above but also because there was a tight exchange of preferences on the left between Labor and Greens which was not matched on the right between LNP, ONP and KAP. 

In short, the biggest winner from the re-creation of ONP was the ALP. In addition to that, Labor was able to hold seats in “the rest of Queensland” by dint of the local campaigns of its various sitting members. In particular, Labor was able to retain all three seats in Townsville as well as its North Queensland seats of Cairns and Barron River. Labor’s win in Maryborough was spectacularly good for its sitting member there, Bruce Saunders.

So why do I say this was “a pretty good result” for the LNP. Essentially it was because the LNP was able to “see off” the ONP. This latter party was able to get a heap of useless votes. 

For Queensland as a whole, the ONP vote was 371,193 or 13.7 per cent. That gave them just one seat, Mirani, taken from Labor. The LNP feared the loss of four seats to the ONP but held them all without trouble. Consequently, next time around Labor will not be able to raise any scare about a possible coalition between LNP and ONP. It will also not be able to attack Campbell Newman. He will be forgotten come 31 October 2020, the date fixed for the next election.

(Malcolm Mackerras is a visiting fellow at the Australian Catholic University’s Canberra Campus.