Prahran and Werribee By-Elections Live

Prahran (Grn vs Lib 12.0%)
Vacancy for resignation of Sam Hibbins (Grn/IND)

Liberal gain from Green/IND (called Sunday 4:20 pm)

Werribee (ALP vs Lib 10.9%)
Vacancy for resignation of Tim Pallas (ALP)
Labor retain (called Thursday 3:30 pm)

----

Distributions: Both preference distributions have been posted and I don't know if there is any scope for further changes so for now it looks like the Liberals have won the primary vote count in Prahran by one vote.  In Prahran, the lead changed a few times through the distribution, and going into the Lupton exclusion the Greens were still ahead. However, that exclusion flowed 63.7% to Liberal.  I have heard from scrutineers that the flow from 1 Lupton votes was 68-32.  

In Werribee, the Liberals gained slightly from the final (Paul Hopper) transfer; here I have heard that the split off the 1 Hopper votes was 60-40 to them but the preferences that had pooled with Hopper from other sources favoured Labor.  What is notable here is that far from the minor candidates preferences pooling with Hopper, he actually went backwards compared to both majors, and got only 27.8% of the preferences at the 3CP stage.  This means that while there was a very large vote for non-majors in Werribee, a lot of those voters were not consistently opposed to the major parties; they just liked someone else more.  The term "double haters" has been thrown around in reference to this result but the distribution doesn't suggest that's what the voters are.  

Friday 6:20 ALP lead in Werribee out to 639, there will be slight changes but I expect it to finish well within 100 of that.  

Friday: Labor has very reasonably claimed victory in Werribee but the result is not official until after the adding of final votes and the distribution of preferences, which I'd expect to happen sometime next week.  In Prahran, it's interesting to note the Liberal primary vote lead over the Greens has come down to seven votes; although the Liberals will win the seat thanks to their superior preference flow, it is still possible they will do so from second on primaries (that would be embarrassing for right-wingers who oppose preferential voting).  

Tony Lupton has been claiming credit for the result in The Australian and suggesting Labor should put the Greens last everywhere.  However not only did his vote not look very Labory in booth terms, but his how to vote card which was orange and said "It's Time For An Independent" was hardly pitched at Labor voters beyond the endorsement from Steve Bracks.  

Thursday 3:30: Recheck primaries have come through without any major changes to Labor's position; this plus the increased lead after postals counted so far makes it clear Labor has retained the seat.   (Update: I understand there are now only provisionals and the last postals to come.)

Thursday 3pm: A report that Labor has done well on primary votes on the late postals and are poised to pull further ahead on 2PP.  (Update: report from Labor that the lead is now 591.)

Thursday midday: Counting of about 2000 Werribee postals has been brought forward to today.  This plus rechecking of booths may put the seat beyond doubt.  

Tuesday: Rechecking in Prahran has not resulted in any significant changes.  Still waiting for rechecked primaries in Werribee.  

Sunday: Where The Lupton Vote Really Came From!

There has been some damage-control spinning from the Greens of a result that is simply very bad.  Among this has been a tweet by leader Ellen Sandell claiming that "Obviously it’s not the result we would have liked but with the unofficial Labor candidate sending their preferences to the Liberals, those Labor preferences have handed the seat to the Liberals this time."  Firstly, Lupton was not a typical unofficial "independent Labor" type candidate but one who'd got involved in some culture war issues and criticised the party.  Secondly, candidates don't send preferences anywhere (it's the lower house) and I'd be surprised if the follow rate for the Lupton card was that high anyway.  Thirdly the evidence is that where Lupton got his votes from tended not to be the ALP support base.  In booth terms, while his vote was pretty flat everywhere, he did worst in the good Labor booths from 2022 and better in the good Liberal booths from 2022 - even though the former had more votes up for grabs care of Labor not running!  Neither result is statistically significant but they do very strongly contradict the idea that the Lupton vote was a large chunk of the former Labor vote - it looks more like an independent conservative sort of thing:

Victorian Labor Kicks The Group Ticket Can Down The Road

(Coverage of Victorian by-elections tonight from 6 pm.  Live page will go up around 5 pm).

Victoria is the last place in Australia where Group Ticket Voting persists in upper house elections.  The system was invented in the 1980s because the Democrats, who are to blame for everything, forced the Hawke Labor government to retain full preferencing in Senate elections.  Because requiring voters to number all the boxes for Senate elections often caused extremely high informal rates, Group Ticket Voting was created as a way to retain full preferencing while cutting the informal rate.  A voter could vote 1 for a party and their party would allocate their preference for them.

Initially this system lacked obvious downsides but its potential for exploitation was obvious as early as the 1987 federal election, where a Nuclear Disarmament candidate with 1.5% of the primary vote was elected.  A series of farcical GTV elections around the country since led to the abolition of the system in NSW, federally, SA and WA leaving only Victoria.  Problems exposed with the system have included:

* parties winning off tiny vote shares defeating much more popular parties when they would not win under any other system

* confusing and deceptive GTV preference allocations that are beyond the understanding of most voters if they tried to follow them

* preference harvesting in which ideologically unrelated parties band together to try to secure election off each others' group ticket preferences

* creation of unnecessary tipping points that should be irrelevant to the contest, making it easier for elections to be voided (eg WA Senate 2013)

* creation of bogus near-100% preference flows between parties when, if asked to choose preferences for themselves, voters spread preferences in a much less concentrated fashion

* corruption of parliamentary voting behaviour, in the form of party votes on electoral reform being influenced by fear of losing the ability to work with Glenn Druery, as stated by Druery himself in the Angry Victorians sting video

* denying voters the ability to direct their own preferences between parties above the line (which they will be used to doing so having done so twice since the last state election) and throwing away their stated preferences and overwrites them with a group ticket vote if they do. 

*confusion between the Victorian system and the Senate system

Poll Roundup: The Long Slow Slide Continues

2PP Aggregate (Last-Election Preferences) 50.6 to Coalition (+0.3 this year)
With One Nation Adjustment 51.1 to Coalition

If polls are accurate, either side could win election "held now", most likely well into minority

We've had quite a few pollsters out early with 2025 being an election year and the nine federal polls so far provide plenty to talk about for a roundup piece.  There have been four weekly Morgans plus one each from Newspoll, YouGov, Freshwater, Resolve and Essential.  All nine have had the Coalition ahead on their headline 2PP measure, and six have had the Coalition ahead on my estimate of respondent preferences.  Nonetheless my own estimate still has the last-election 2PP rather close.  

Voting Intention Fine Details

This section has a little more wonky detail than usual, mainly to explain the ins and outs of why my aggregate is better for Labor than the headline preferences in the current batch of polls, but also to explain why it could easily be a bit worse.  While I have Labor on 49.4 by last-election preferences, it could arguably be 49.2.  

Going The Distance: Federal Seats That Do And Don't Make It To The Final Two

Recently I had this question on the website formerly known as Twitter.  

I thought this was a very interesting question because while whether a seat comes down to the final two or not doesn't have much to do with how close the seat is, it's something we don't think about much in federal elections.  Every vote is thrown by the AEC to the final two candidates, but in some cases it doesn't reach the final two until after the seat has already been decided because a candidate has crossed 50%.  Indeed, in a rapidly reducing number of seats the contest is decided on primary votes and the preference throw is entirely academic to the result.

Contests don't get thought about in these terms so much because every vote is present in a notional final two soon after it is counted via the notional two-candidate preferred.  This gives a single indication of the margin of victory.  It obscures the nuance that a candidate who finishes on 54% of the final two-candidate preferred might have even crossed 50% before the last exclusion then done badly on the last lot of preferences, while another candidate might have needed a trickle of preferences to get over the line yet got so many preferences that they finish up with a 2CP of 68%.

Prahran and Werribee By-Elections

In a nice appertiser for the federal election year, Victoria has a couple of unusually interesting state by-elections coming up on February 8, which I intend to cover on the night.  One is especially interesting because it's a more or less stock-standard test for the theory that the a government is in trouble, the other because of its uniqueness.  This is a preview post for these by-elections where I look at some factors that might affect the results; I will add notes on polling if I see any.  

Prahran (Green vs Lib 12.0%, 3CP Green vs ALP 5.8%)
ALP not contesting
Vacancy for: resignation of Sam Hibbins (Grn/IND) following personal scandal

Prahran is a unique seat to begin with, and this by-election is even more of an oddity because of the decision of the Allan Labor government to not turn the rock over.  Labor cited the fact that they had not won the seat since 2006 and didn't need it as pretexts for not having a go.  Winning would have been challenging because the standard anti-government by-election factor would have made it hard for them to get over the Greens even with the loss of Hibbins' personal vote.  On the other hand with the government currently not polling well there could have been fear of a really bad result (such as being jumped by some indie and finishing fourth).  If Labor could have won the by-election they may well have been in a better position to hold off the Greens at the next general election, but unlikely to happen and not to be.

2024 Site Review

Bring Back Universal Analytics Again!

As in my 2023 review, I again open this annual site review by expressing my disgust at Google stuffing with the continuity of my previous site stats by replacing Universal Analytics with GA4, which is not only apparently worse and different but also very hard for me to understand.   Anyway, again, on with the show as best I can ...

The dotted top line of the graph below shows activity in 2024 in the form of user numbers per week.  There is a big spike on the left caused by the Tasmanian election buildup and post-count, and a substantial spike on the right for the overlapping ACT and Queensland elections.  Other spikes correspond to the Tasmanian Legislative Council counts in May and the NT election in late August.  

Site engagement as measured in total events was up 84% on 2023, making 2024 as best I can determine the second-biggest year for the site after 2022.  That said, the new Analytics version seems to show high levels of site activity relative to activity on specific posts when compared to the old one, so I am not sure that is reliable.

In 2024 I released 79 articles, up 13 on 2023.  This included 27 articles about the snap Tasmanian election. There were also six articles about the Queensland election, five for the 2024 Tasmanian Legislative Council election and four federal poll roundups.  

Twitter Vs Bluesky And The Psephosphere

Last year's introduction will do just as well for this one.  It's an almost annual tradition on this site to release something every Christmas Day.  Click the Xmas tag for previous random examples.  Why do I do this?  Partly it's a present for those who like Christmas, which in a very low-key non-religious fashion includes your host, but it's also a present for those who don't want to deal with this particular Christmas or even generally cannot stand Christmas and just wish it was a normal day when normal things happened.  And what could be more normal than niche meta-psephology?  Therefore, the campaign against compulsory Christmasing brings you ... whatever this is.  

As the owner of X, which I still call Twitter, continues to run it into the ground both by making almost invariably bad changes to the website and by generally being himself, interest in alternative microblogging websites has grown.  In June 2023, with Twitter appearing in some danger of permanent technical collapse, I signed up accounts on Bluesky, Mastodon and Threads to cater for those fleeing the mothership, and as an insurance in case Twitter died or downgraded its features to a point where I wanted to leave it.  Microblogging sites are very important to my work, both as a place to obtain news about electoral matters but also as a place to make people aware of newly released posts on here (given how irregularly I post them.)

Until recently my activity on Bluesky, Mastodon and Threads has been mostly a secondary service with links to articles on this site posted; sometimes I will crosspost a debunking of some particularly bad poll-shaped object, and usually I will reply to questions when I can.  However this changed in the wake of the 2024 US election, when there was a mass exodus of Twitter users to BlueSky as a result of Elon Musk's active support for the return of Donald Trump.  Over the coming weeks my total Twitter following dropped from about 16900 to 16400 as some of the refugees deleted their Twitter accounts.  At the same time my Bluesky following rose within a week from around 950 to over 2000, even though I was not initially posting that much there, and it has now reached nearly 4000.  Clearly plenty of account holders have become active on Bluesky without deleting Twitter, though some may have stopped posting actively on the latter.

2024 Federal Polling Year In Review

 2PP Aggregate Average For 2024: 50.9 To Labor (-3.9)
Labor lost aggregate lead late in the year

It's the time of the year when most busy pollsters take a few weeks off and I bring out an annual feature, a review of the year in federal polling.  Click here for last year's edition and for articles back to 2014 click on the "annual poll review" tab at the bottom of this one.  As usual if any late polls come out I will edit this article to update the relevant numbers.

2024 was another strong year in results terms for the Australian polling industry.  Pollsters came out in good numbers for the early 2024 Tasmanian election and did pretty well in a very hard to poll election, although the lobby group(s) that commissioned two Freshwater polls contemptibly failed to ensure the release of definitive results of either, leaving poll-watchers to play jigsaw puzzles with incomplete media reports. Polling for the Queensland election was mostly excellent though no one pollster nailed the result, and a mini-cluster of close-ish polls at the end led to some misreads of what in the end was not a close election.   Despite this there are storm clouds about in federal polling in the form of inadequate transparency from several pollsters and a somewhat suspicious level of clustering of results, especially at a stage in the cycle where that doesn't usually happen.  (More on the latter later).  US polling this year wasn't as bad as 2020, but didn't quite get the real story in part because of the latter issue.

Media Fail Again Over MRP Reporting

Redbridge/Accent MRP model projects Coalition would be likely to have the most seats if the federal election was held now.  

It does not say Coalition would be likely to have a majority.

It does not predict the result of the election.

---------------------------------------------------------

I thought an article was in order to unpack some of what is going on with the recent Redbridge/Accent MRP model and the woeful reporting of it by pretty much every outlet that has so far mentioned it.

MRP models (stands for Multilevel Regression with Poststratification) are not the easiest to explain to lay readers at the best of times, but what we have seen from several media sources reporting on this one goes beyond understandable confusion and into the realms of reckless innumerate false reporting.

What a MRP model does is to build a picture of how certain types of seats are likely to vote based on small samples of all 150 electorates.  Although each seat's sample is uselessly tiny by itself, by assuming that seats that resemble each other in ways that affect voting intention will vote similarly, one can smooth out a lot of the rough edges in the sampling, and samples of a few dozen voters per electorate can build a model that's about as good on a seat-by-seat basis as if those samples were actually a few hundred.  That still isn't very good on a seat-by-seat basis, but on a nationwide basis, the model could capture some general trends about the kinds of seats where each party is likely to be doing well or badly, and about how a party might be going in converting vote share to expected seats.