There’s been a lot of discussion comparing the recent City of Brisbane council election to the 2022 federal election. Both elections saw a high Greens vote. At the federal election, the Greens won three seats in the inner city of Brisbane, something which exceeded expectations. Expectations were higher at the council election, with talk of the Greens winning up to six wards. The Greens did gain a swing, and won a second ward, but the eventual results didn’t quite meet expectations.
There are quite a few differences between the two elections which plays into the different outcomes – a lower Labor vote, a higher LNP vote, and a different preferential voting system, all of which helped the LNP in a relative contest against Labor and the Greens.
So I thought I would try to compare like with like – the primary vote in the two elections.
I was able to convert results of the 2022 federal election to the 26 Brisbane City Council wards, which then allows me to compare the primary vote at the two elections. To start with, let’s look at the council-wide change in vote.
Just to confirm this is just counting the federal 2022 results within the City of Brisbane. In the case of electorates that overlap the border, it’s only counting those SA1s contained in the City.
The Liberal National Party vote was about 11.5% higher at the council election, while the Labor vote was 5.9% lower.
The Greens vote was almost the same at the two elections. Indeed it was actually slightly higher at the council election. The proportionally lower seat result was not due to a lower primary vote, at least across the whole city.
Also bear in mind that council ballot papers were much smaller than for federal elections, so the ‘other’ vote was much lower, which explains why the change in vote doesn’t add up to zero.
Then next up I have made some maps showing the change in vote for each party in each ward from 2022 to 2024.
For the Greens map, I’ve overlaid the three federal electorates they won in 2022 in red lines.
The Greens vote went up in fourteen out of 26 wards, which is in line with a slight increase citywide.
The biggest changes tended to be in inner city wards targeted by the Greens, with the primary vote 5-10% higher in Central, Paddington, The Gabba and Walter Taylor, and a smaller increase in Coorparoo.
There were also significantly higher Greens votes in Moorooka and Forest Lake, while the Greens vote was significantly higher in a string of wards on the northern edge of the city.
If you overlay the federal electorates, the biggest increases appear to have been in Ryan and Brisbane, the two federal seats the Greens won by the smallest margins. Griffith is more of a mixed bag, with the Greens doing significantly worse in Morningside. This might be explained by some campaign effects and incumbency, with a Labor local councillor and the Greens not focusing on the ward, but it’s also worth noting that the Greens had a 9.9% swing on primary votes in Morningside, their fourth best result. It just wasn’t as good as the Greens vote in that area in 2022.
I also made maps for the LNP and Labor vote but they are less interesting. The Labor vote was lower in most areas, and the LNP vote was higher in most areas.
So what does this tell us about the different outcomes in 2022 compared to 2024. Firstly, it shows that it was not due to a differential primary vote – the Greens had a big uptick in their vote in all of the wards where they came close to winning.
They lost those wards due to a combination of different shares of the vote for the other parties, and optional preferential voting. The LNP started a lot further ahead than they had in 2022, and Labor was much further back, thus having less to offer. OPV also meant that those Labor preferences were less helpful in reining in the LNP.
It also suggests that the Greens may be having some success in increasing their vote in their inner-city heartlands now that they have incumbent federal MPs, although this doesn’t account for differences in voting patterns between the two levels of government. This is obviously a major factor in support for Labor and the LNP.
It is important that this not be taken as a prediction – it doesn’t mean the Greens will do better in 2025 in areas where their 2024 vote was higher. But it does debunk the suggestion that the Greens not winning as many wards as they did electorates in 2022 was due to the decision of voters to not vote for the Greens. More people voted for the Greens in 2024 than in 2022 – but the other voters tended to vote in a way that was less helpful for the Greens, and were strongly encouraged to do so by the OPV system.
How does one explain the increase in the Greens vote in Forest Lake ward?
Also, is the presence of independence and minor parties sufficient to explain the gap in the Greens primary vote between Forest Lake ward and Inala district (in the by-election)?
I want to add a factor that was once standard knowledge of Queensland voting behaviour: Queenslanders were always able to differentiate between the different levels of government. The fact they supported a federal Labor party didn’t mean they voted for Labor at a local government level. It depended how they valued the work of that level of government, whoever was in power. It was a distinctively Queensland characteristic. These days Queensland culture has been diluted by high levels of immigration. The culture may have changed.
Interesting about Hamilton Ward, Labor’s vote down and Greens went up but this still underperfomed the Federal results. I guess the LNP vote held up much better in many affluent seats.
I will start with my usual disclaimer that I am not a psephologist, but one thing I have noticed in a lot of commentary prior to the recent LG election – a lot of people seem to be relying on overlying the federal election results and then applying that to the wards expecting Federal Green votes to automatically transfer over to LG Green voters.
Sure there will be a degree of “rusted on” voters for any party, but in my mind a vote for Greens in the Federal election doesn’t happen to mean that it will automatically transfer over to State or Local Government elections. You have to give voters SOME credit for being able to differentiate between different levels of governments and the different roles, responsibilities and issues they will address. For example I expect the Greens to have a stronger vote in the upcoming State Election because their main issue (housing and renting) is mainly a State issue because this level of government sets the legislation and have control of the institutions and actual actually DICTATES to LG what they must do (e.g. housing targets in certain areas). So if they keep pressing this issue, they can make a more forceful argument for a Green vote at a State level.
Speaking to the Greens candidate in the Moorooka Ward prior to the LG election she made the point that much of their resources come from members already sitting in parliament (Federal and State). So it is only natural that they concentrate on seats they they already hold. She said she was only getting a small amount of money for her campaign. Whereas both major parties have enough cash and resources to spread themselves more widely.
Much of inner Brisbane currently has a federal Greens MP, a state Labor MP, and an LNP councillor.
I’m not sure this is the best methodology to use for that map. If I understand you correctly, you’re comparing the primary vote across a whole federal electorate to the primary vote for the wards/SA1s that fall in those federal electorates.
So, for example, Griffith had a Greens primary vote of ~35% at the federal election and the wards that fall within Griffith are being compared to that. Wards that had a higher Greens primary in 2024 are green (and lower in red).
But I think shading it this way gives the impression that the Greens’ support grew in the inner city and declined in the middle/outer suburbs between 2022 and 2024. You’re comparing booth results to the primary vote for the whole federal electorate, which is effectively an average across a much larger area. There are always booths that do better for the Greens (especially in West End) and booths that don’t do as well (typically further out). And that’s what the ward results show. The difference between ward results and federal results is probably more to do with ward results being more granular than looking at a whole federal electorate at once.
If you were to plot federal results by booth and compare those to the nearest booth results of the wards, I’m guessing the +15% and -15% results would mostly disappear. Alternatively, if you were to bundle all the wards together, the average result would probably be close to the federal result. (Which is basically what you did in the graph.)
This is all to say that I agree with your overall point: the different voting systems used have a larger impact on the results than the idea that the Greens underperformed in the LGA election.
No, I’m not comparing ward results to whole federal electorates. I have a dataset of estimated federal vote by SA1, which has been created by using the AEC’s dataset telling us how many people from each SA1 voted at each booth or method.
So I then found the dataset matching SA1s to the BCC wards that I created in 2019 for the last ward redistribution and added up those votes.
So it’s not perfect, but it’s very much adjusted. Different wards within each electorate have different federal votes, which you can see if you click on the map.
For example, I have estimated the Greens federal vote at 39% in the Gabba, 34% in Coorparoo and 32% in Morningside, even though they are all entirely contained within Griffith.
Ah, that makes sense. I didn’t pick that up from the description in the article.
Nicholas, I think the result in Forest Lake is down to a couple of factors:
– the local Labor branch being disappointed with sitting councillor Charles Strunk for firing a well-liked staffer, which may have caused both a reduction in Strunk’s campaign activity and a protest vote for the Greens amongst some usual Labor voters.
– the local Vietnamese Australian community being disappointed in Labor not picking a member of their community to represent them at the Inala by-election. The Greens had a Vietnamese Australian candidate in Forest Lake, Vi Nguyen. The LNP got a similar swing to them in the Inala by-election after they ran a Vietnamese Australian candidate, Trang Yen. You might call this the Fowler effect.
@ Wilson
The Greens got a good increase in Morooka as well. Could that be due to increased support among the Muslim community due to Palestine?
Nimalan, while I wouldn’t definitively rule that out, I would consider it less likely to be the main reason for the Greens vote increase in Moorooka. Despite Moorooka having many African shops and restaurants as well as a mosque, demographically I don’t think the area has a particularly large share of Muslim residents anymore, as rising house prices have forced many of them further out to more affordable areas such as Logan and Ipswich. Many of the people who have moved into the area in the same time are middle class younger people who have been priced out of suburbs closer to the city, but can afford Moorooka house prices.
Fair Point Wilson
Agree Logan especially is seeing a growth in the Muslim community. I am keen to see how Wills will be at the next federal election as it is almost 10% Muslim and concentrated very much in the Working Class North to see if this helps the Greens there.
We are still waiting for seats to be declared and final figures to be made available for the Brisbane City Council election. I understand the use of Federal primary votes as a comparison, but its value depends upon a few assumptions we need to hold on mind. One assumption is that the last Federal election serves as a good benchmark. However, there was a swing in that election where a long running LNP government was decisively voted out and it is likely that a number of protest votes went to the Greens. It may easily prove to have been the high tide mark for the Greens. Secondly, Queensland voters have always been known for separating out different levels of government in choosing which party to support. In that case, making these sort of comparisons between federal state and local government elections may not be as valuable as elsewhere in Oz. I think the Greens were reading from the same prayer book as you, Ben, when they claimed seats on election night that they went on to lose. Looking at the ABC’s prediction of final figures for the election we see an increase in the LNP vote and a larger increase in the Greens – but the Greens increase is completely at the expense of the ALP. That is in my opinion the most noteworthy trend in this election. Waiting for final figures but the image in my head is of a vampirical Greens political party feeding on a sleeping ALP.
Lynne, I don’t know what you think I’m doing. I’m just measuring how high the Greens tide was at this election and comparing it to the previous high tide. And the result is that they did better at this election.
I also don’t know what this has to do with claiming victory in seats they go on to lose. I’m not saying anything about wins or losses.
Interesting to think that the Labor + Green primary votes indicates a progressive majority in Brisbane but the council is dominated easily/totally by the LNP. Thinking ahead, obviously the LNP State government likely to be elected in October will swing back to optional preferential voting at State level for 2028. It could mean slim picking for progressives for the foreseeable future in Queensland. Perhaps I should invest in companies manufacturing pink tracksuits?
Roger, I believe it’s too early to predict the complete downfall of Labor in Brisbane and Queensland overall. After the 2012 wipeout, many analysts believed Labor would be consigned to decades out of power but they rebounded quickly after just one term.
Obviously it depends on the lnp making huge missteps but even if they end up being a strong and competent force, they will eventually become stale and worn out just like the nsw coalition which finally ended up not having enough juice to get over the line for a record 4th consecutive term.
Even the Tasmanian Liberals who were also seeking a record 4th consecutive term like the coalition in nsw also suffered a significant loss of support and are now forced into minority.
@nimalan:
My experience of Muslims is that they’re a strong Labor Demographic because Muslims understand Machine Politics.
Greens support for Gaza, not saying it’s opportunistic or cynical, but it plays well to non Muslim female voters.
Annerley and the part of Tarragindi along Ekibin Rd contain a lot of Princess Alexandra Hospital Workers, they are nearly as strong Greens Voters as Labor Voters. The area is mostly stand alone timber houses and a little run down. These are your strong Greens voters, bought the house when houses were cheap, lives close to work, living in a run down house on land worth $1m.+
Higher up in Tarragindi near Chardons Corner is a very pretty part of Brisbane, LNP do well here and Greens drop down to their average.Further out into Rocklea and Coopers Plains Labor does very well and the Greens are ahead of the LNP.
Here’s one that voted 55% Green, there were only 20 votes. Church St Wooloongabba, a very private street 5 minutes walk from the P.A. https://results.aec.gov.au/27966/Website/HousePollingPlaceFirstPrefs-27966-6332.htm
Still five seats to be fully counted, but I’m interested in whether Jonathan Sriranganathan was a benefit or a liability for the Greens vote.
Interesting that the ECQ appears to not be doing a distribution of preferences if the candidate won on the primary vote. They’ve also declared wards like Marchant where the preference count is ongoing.
@ Gympie,
Regarding the Muslim vote, there has been suggestions after the 2022 election that the LNP should target this demographic based on social conservatism especially on LGBT issues. There was also less geopolitical focus on the middle east. The seat of Calwell had the biggest anti-labor swing in the country. I was skeptical about Outer Melbourne as i feel that was isolated to the lockdown backlash. However, one area of concern about Muslim voters potentially drifting away can be found in parts of Liverpool LGA with Labor under performing in Werriwa and the state seat of Liverpool. I think Ned Mannoun may have helped here. In the UK we can see George Galloway picking up Muslim voters angry about Gaza in the Rochdale by election and the recent Michigan Democratic primary.
@Mark, I expect they will still do a distribution, but they don’t need to finish the counting before they declare the result.
The AEC works the same way – once it’s mathematically certain who is won, they declare, but it doesn’t mean they stop counting.
Ben you’ve asked: “ I don’t know what you think I’m doing. I’m just measuring how high the Greens tide was at this election and comparing it to the previous high tide. And the result is that they did better at this election.” To me, the significant point is that the growth of the Green party vote was only at the expense of the ALP. The Green party has bought your idea their vote is surging hook line and sinker. That’s why they saw victory mirages on election night. Interestingly, having the Green Party rather than the ALP run second is an advantage to the LNP. My observation is that the Labor preferences flow more strongly to the LNP than the Greens preferences do. I am looking forward to seeing the preference figures when the ECQ gets around to providing them. Spectacularly slow this election.
Yeah I’ve already calculated preference flows based on the notional 2CP counts on the night, and while both Labor and Greens preferences to each other increased in 2024, Greens preferences to Labor are definitely stronger than the other way around (but only by a bit).
I suspect even in 2022 that the Greens mostly gained votes from Labor. To the extent that the LNP lost ground, it was probably a parallel process of Labor gaining from LNP (as happened all over Australia) and Greens gaining from Labor. This time around, only the latter happened. And you need both to win close races.
There are three seats where preference distribution has been stated The Gabba where the new Green incumbent obtained 1.5% less after distribution than Sri, and I note that 4% more Labour voters exercised a second preference, two perecent to Liberal National and 2% to Green.
Enoggera where Greens displaced Labour for second and obtained 1.2% less than Labour in 2020, but obtained just over 50% of Labour second preferences. Green final preference jumped 30.7%.
Pullenvale where Green remained second but dropped 4.4% in a final preference situation where the Independent vote dropped and the LNP first preference vote went from 41.7% in 2020 to 53.8% in 2024.
Thanks for this analysis, Ben. Immediately after the BCC election, I saw lot of commentary to the effect that Bates and Watson-Brown might be in trouble at the next fed, but if this shows anything I think it’s that voters in Ryan and Brisbane aren’t feeling any buyers remorse for voting Green.
As for Griffith, you mention Morningside, and as a local I can confirm that the Labor incumbent ran a good campaign and the Greens didn’t /really/ bother. A few events and flyers here and there but nothing like in their main targets. I actually heard way more from Chandler-Mather than their Council candidate in that period. So it doesn’t surprise me that Morningside went against the grain. If that Ward is set aside, does the rest of Griffith follow the pattern In Ryan/Brisbane?
A bit of Doboy is in Griffith and that has a similar trend (although I can’t say if it was the specific parts of Doboy in Griffith). And Coorparoo didn’t give the Greens as much of a boost as the other target wards, but generally yes.
It’s worth noting the Greens still got a big swing in Morningside. They polled ~15% there in 2020, ~25% in 2024 and ~35% in 2022.
Thanks – not surprised by that swing. If they ‘didn’t really bother’ in 2024, they ran completely dead in 2020. Unappealing, invisible candidate. Don’t remember getting any flyers from them at all.
I am surprised by Coorpooroo though. I’m not really that in touch with the community, but I had the impression that Kath Angus was gonna walk it in.
Obviously the Greens got a decent swing in 2024 in Coorparoo compared to 2020. But compared to Max Chandler-Mather in 2022, she only polled 0.6% more. You can see this yourself on the map above (although you have to turn off the federal seat overlay to be able to click on the wards in that area.
The main explanation for the Greens winning less wards other than OPV is that many state and even federal Labor voters vote for the LNP on the local level. BCC Labor are obviously more progressive than QLD Labor and even federal Labor while the BCC LNP are more moderate than the QLD LNP or the federal Coalition. In most wards the BCC LNP vote is higher than the federal Coalition vote by over 10%.
An example of this is Nudgee. Nudgee is a suburb in the safe Labor seat of Nudgee on the state level and the safe Labor seat of Lilley on the federal level. However, Nudgee is also located in Northgate, an ultra-marginal BCC ward currently held by the LNP with a tiny margin.
@ NP
I agree much of the reason why Greens underperformed the Federal result is due to BCC LNP being more moderate than their federal counterparts. It is the same in NSW where the NSW Liberals are moderate so they did much better in Teal areas.
@Nimalan, I think it is more to do with Queensland having more right-leaning demographics that still affects even in Brisbane’s inner-city. Afterall, Ryan still has a sizable No vote (47% No) indicating conservative politics still perform better in Brisbane as a whole than a equivalent area in Melbourne for example.
Marh, I don’t think you can directly compare suburbs between capital cities (especially Melbourne to Brisbane).
A large chunk of Ryan (almost half) is actually comprised of relatively sparse suburbs and is considered semi-rural in nature, almost like the Dandenong Ranges of Melbourne (places such as Lilydale and Belgrave, which are naturally conservative leaning anyway). Only the closest suburbs bounded by Indooroopilly and Mt Coot-Tha resemble the inner-city parts of Melbourne which are strongly progressive .
Brisbane is a poor comparison to use when trying to overlap federal seats to either Sydney or Melbourne. Due to its small size (in population terms), you transition very quickly between demographic areas. Also, there is less diversity with the city as a whole having greater proportion of those from European or Anglo (ethnically white) backgrounds and not as many CALD minorities.
@Yoh An, even adjusting for demographics, Brisbane is still more Conservative and/or No Vote for The Voice leaning than Melbourne equivalent areas. Compare Ipswich and Frankston. Also Lilydale, Belgrave and semi-rural parts of Ryan all completely different areas. Lilydale is a middle-class suburb with lower percentage of University Qualifications but votes more favourably to the Libs which could literally very similar to many outer suburbs of Brisbane, Belgrave isnt even a conservative leaning area, its a progressive town on the bottom of the Dandenong Ranges and Kenmore area is a wealthy area mainly fiscally conservatives.
@ Marh
While i agree that QLD does have underlying right-leaning demographics compared to the rest of the nation, i believe that is due to it being very decentralized and the majority of population is outside Brisbane. Another factor is a lack of manufacturing industry in Brisbane compared to Melbourne. However, i do agree with Yoh An, that most Brisbane electorates cover at least 2 demographic groups. Even Rankin, cannot easily be compared to Working class ethnic seat such as Calwell, Blaxland etc eventhough quite a lot of Rankin fits that profile hence Rankin has a lower 2PP than the seats i mentioned and is more socially mixed.
@Nimalan, I do agree due to small population however a Rankin-like seat still does happen. in Sydney (Parramatta e.g.)and Melbourne (Bruce at 1996-2016 boundaries e.g.).I think it has to do with Brisbane social-economic areas aren’t bounded with a particular metropolitan region like you see in Sydney, Melbourne and even Adelaide.
Agree Marh, Brisbane doesn’t have the same sort of socio-economic areas that are featured in Sydney and Melbourne. One example is industrialised suburbs with unionised workers – examples would be Yennora for Sydney, Dandenong for Melbourne and Elizabeth/Northern Adelaide. A comparable area for Brisbane may be Moorooka/Rocklea but whilst that is an industrialised area, the workers there are not really unionised in a manner that is characteristic of the other localities.
Agree Marh and Yoh An
In Melbourne or Sydney, areas like North Shore, Inner East, Bayside etc are a Euphemism for a particular social class which is associated with a region while you don’t really get that in Brisbane and suburbs that are rich and poor can be close together and drawn in the same seat. As you mentioned Adelaide and even Perth show that with particular regions associated with voting patterns and social class. Good point about a Rankin like seat being the Old Bruce, Parramatta it will be interesting if a super safe Labor seat can be drawn by combing Inala, Woodridge, Rocklea etc that will have a margin like Calwell or Blaxland.
Nimalan, if Queensland was a state in the US that type of contorted district combining disparate areas/suburbs would be one that would be easily drawn under an LNP gerrymander to pack as many Labor voters into one seat whilst leaving the others across Brisbane to be slightly conservative in nature (albeit marginal in nature).
If a US style gerrymandering happened Sydney and Melbourne, I calculated most seats would have a weird shape (both Labor and Libs biased) given the large amount of marginal/relative marginal voting booths