For this post, I’m not going to dwell too much on individual races and where they are up to, but instead look at broader trends. I’ll do another post today or tomorrow about the close races and what we’re waiting to see.
There were quite a few votes that weren’t counted last night. Almost 600,000 votes were counted for the lord mayoral election, but barely 450,000 primary votes were counted for the council election, and only about 60% of those votes had preferences distributed.
Overall the story is simple. The LNP’s vote largely stood still, while there was a substantial swing from Labor to the Greens. The Labor vote is still larger than the Greens vote, but not by a wide margin.
At the time of writing, Adrian Schrinner’s primary vote is up by just 0.2%, while the LNP council team is up by 0.7%. The swing from Labor to Greens was 4.5% on the mayoral ballot and about 6% on the council ballot.
The LNP is not winning an enormous share of the vote – Schrinner is on 48% of the primary vote and 55.8% after preferences – but with a single-member system with optional preferential voting it’s enough to win in a landslide when facing two opponents who split the anti-LNP vote fairly evenly.
There is a clear geographic polarisation of the centre-left vote, with the Greens polling better in the inner city and western suburbs, and Labor still being dominant in an outer suburban ring.
This first map shows which of the centre-left parties is polling higher in each ward, and can be toggled to show how much of the combined Labor+Greens vote was for the higher-polling party.
Labor is very much in the position of a minor party in some of those inner city wards. The party polled under 20% in Central, Paddington, Walter Taylor, The Gap, Pullenvale, Tennyson and Coorparoo.
This table shows the primary vote for the twelve wards in the inner/western suburbs, and the fourteen wards in the remainder of the council.
Region
LNP
ALP
GRN
Inner & Western Suburbs
44.0%
20.1%
30.8%
Outer Suburbs
46.7%
34.5%
17.7%
The LNP does slightly better in the outer suburbs, but face very different opponents. The Greens poll a vote 50% higher than Labor in the inner city, while Labor’s vote is twice the Greens vote in the outer suburbs.
Labor reached the 2CP in all fourteen outer suburban wards, but just two inner city wards (Holland Park and Morningside). An independent has been re-elected in Tennyson, but the other nine inner city wards are LNP vs Greens contests.
Of course there is not a hard-and-fast boundary between Greensland and Laborland. In eighteen wards, the higher-polling centre-left party polled less than double the vote for the other party. The parties are basically tied in Jamboree, and are also very close in Holland Park and Hamilton.
The use of optional preferential voting becomes more of an issue in wards where Labor and Greens both poll a substantial share of the vote. I have previously posted about how I believe three wards would have flipped from LNP to either Labor or Greens if compulsory preferential voting was used. I’m sure we’ll find something similar once we have a more complete dataset for this election, indeed there may be even more. There are a lot of close wards.
Optional preferential voting is often talked about as this force of nature, almost equivalent to first past the post. But some preferences do flow under OPV. While FPTP removes choice from a voter by forcing them to choose between their favourite candidate and the candidate with the best chance of winning, OPV doesn’t do that. And we do see rates of preferencing change depending on the political context.
Based on the early data, it looks like there’s been a significant uptick in rates of preferences flowing from Labor to Greens and vice versa.
This first table compares how preferences from all minor candidates flowed to the final candidates at the 2020 Brisbane City election to the flows last night. I excluded any primary votes in booths where a preference count has not been concluded, so it is definitely a partial dataset. And this also includes preferences from other minor candidates, but there aren’t many of those.
LNP
ALP
GRN
EXH
ALP in top two, 2020
10.2%
46.3%
43.5%
ALP in top two, 2024
11.4%
57.4%
31.2%
Greens in top two, 2020
13.5%
35.4%
51.1%
Greens in top two, 2024
17.0%
49.2%
33.8%
Generally Labor preferences don’t flow to the Greens as strongly as vice versa, but the exhaust rates in both scenarios have dropped substantially. While half of all preferences exhausted in Greens contests in 2020, that rate is barely one third in 2024.
The number of votes flowing to the LNP did go up, but by much less than Labor or the Greens. Even if the ratio of preferences flowing to the LNP stays steady, it’s still more helpful to their opponents to have less votes exhausting.
But was that just a relic of 2020? I wanted to look over a longer period, but council elections are messier, with different ranges of candidates, wards with no Greens candidate, and changes in who made the 2CP. This isn’t true for the lord mayoral election. This next chart shows how preferences from other candidates (predominantly the Greens) split between LNP, Labor and exhausted votes at the last six elections, plus last night. The lord mayoral election also has the advantage of having a more complete dataset than the council election.
It looks like there has been a long-term trend of Labor gaining more preferences, which probably reflects the increasing vote share of the Greens (whereas most of the “others” vote in 2000 was from independents), but still the 2024 data looks quite different to 2016 and 2020.
This probably explains why the ABC and Poll Bludger were having a lot of trouble projecting results last night. With the ECQ taking a long time to count (partly explained by counting the lord mayoral election first, but it seems like there are other issues), they were relying on assumptions about how preferences based on past elections. If those preference flows change, those assumptions will be wrong.
The Brisbane election is a brilliant example of how electoral system design have massive impacts on outcomes.
The LNP is not overwhelmingly popular – the LNP hasn’t won a majority of the vote at either the 2020 or 2024 elections, but look set to win a significant majority in addition to the mayoralty. There is no single centre-left party that can rival them in popularity, but Labor and the Greens did poll a majority of the council vote between them.
It is true that the result would probably do a better job of reflecting how people voted under CPV than OPV, but I don’t think that should be the end of the story. Any single-member electorate system will do a very poor job of representing a community that is voting the way we saw in Brisbane last night.
It doesn’t have to be that way. New South Wales, Tasmania, the Northern Territory and to a lesser extent South Australia and regional Victoria have proportional electoral systems that do a better job of representing voters. Even three-member wards would produce much better results and make the question of whether voters are forced to mark preferences much less important.
The Queensland government has now been in power for nine years and has totally failed to tackle the poor state of Queensland local government electoral systems. The block vote system used for undivided councils is an embarrassment, and the single-member wards used in the south east aren’t much better. Queenslanders should demand better.