Aston By-Election 2023

ASTON (VIC, LIB 2.81%)  April 1 by-election

Liberal to be announced vs Mary Doyle (ALP) and others
Cause of by-election: Resignation of Alan Tudge

Outlook: On historical patterns Liberals should retain - but it may not be easy

Yesterday what looks like a quiet electoral year was enlivened just a little by the news that there will be a by-election for the Liberal seat of Aston following the resignation of Alan Tudge, MP since 2010 and Minister between 2016-2022.  Tudge was once a minor star of the Coalition government but his rise was tarnished from late 2020 onwards by an extramarital affair with staffer Rachelle Miller, who accused him of harassment and abuse.  Tudge denies the accusations, which led to a substantial payout.  In recent weeks Tudge has been under the spotlight for his actions surrounding the illegal, disgraceful and barbaric "robodebt" scheme as Social Services Minister in 2016-7.  Tudge has cited family and health reasons in his resignation, having flagged his intentions with Opposition Leader Peter Dutton in early January after his father passed away. 

This thread will follow the by-election campaign though I will lose interest in it and stop updating the thread if Labor doesn't run and there is no other serious challenger.  [Update 17/2 Labor is running and has re-endorsed Mary Doyle.]

Governments normally get a whacking in by-elections, so why should this one be any different?  Firstly it is an Opposition vacancy, so the Liberals will not have whatever personal vote Tudge had available to them at this election.  As noted in my preview for the 2020 Eden-Monaro by-election, where there was a small swing to the government of the day, the average swing in Opposition vacancy by-elections that both parties contest is about 1.2% to the Opposition.  (There's no evidence that this is lower in first-term by-elections.)  

Secondly, the Albanese government is currently polling very well indeed, and there is a reasonable correlation between federal government polling and by-election results.  In opposition seats, the historic regression suggests about a 1% swing to the government, though that is heavily influenced by one data point (ACT 1970) without which the projected swing would be about zero.  All that said, the 2008 Gippsland by-election was a fizzer on that front with a 6.1% swing to the Nationals, kickstarting the career of Darren Chester.  As noted by William Bowe, Labor was also unable to turn the early Hawke honeymoon into any seat gains.  In general, where substantial by-election swings to governments have occurred, the Coalition is the beneficiary.

Famously, there have been no government gains from the opposition in federal by-elections since Hugh Mahon was unjustly expelled from the parliament then defeated in 1920.  But that's not to say it's as implausible as it sounds - the cases of swings to the government in such elections just haven't ever again quite lined up with the margins needed for a seat to fall.  Someday it is going to happen again.

There is not much risk for Labor in running - even a swing to Labor would produce some short-lived embarrassment for the Opposition, while a swing to the Coalition could be written off as a product of Tudge's unpopularity or just the usual by-election swing against governments.  

Aston form guide

Aston, in Melbourne's eastern suburbs, was created in the 1984 expansion.  It was at first won by Labor's John Saunderson (previously MP for Deakin) but Saunderson's 1984 and 1987 wins are the only times Labor has won the seat and the only times Labor has outperformed the national 2PP in the seat. The shift away from Labor is credited to demographic factors.

Since then it has been continually Liberal, at times fairly safely so and at times marginal.  It has normally run about five points above the national Liberal 2PP, with notable exceptions in 2004 (10.5% above) and 2016-9 (8.2% and 8.6% above) on the upside.  On the downside it was only 2.8% above the national Liberal 2PP in 2007 and 1.9% above in 2010 - in the latter case it was vacant and Labor performed well in Victoria generally with Julia Gillard as Prime Minister.

Aston's previous by-election in July 2001 was significant.  The Howard government had been in major polling trouble earlier in the year but its stocks were fast improving.  Even a routine by-election swing for a government vacancy would have handed Aston to Labor, but the Liberal Party just held on.  Even before the Tampa and 9/11 attacks, this cemented doubts about whether Labor was still on track for victory.

Something interesting happened in the seat in 2022.  There was a large 2PP swing to Labor of 7.32%.  This was somewhat larger than the swing in adjacent Chisholm, and significantly larger than other nearby divisions.  However, the 2PP swing in the Senate was almost exactly the same, and the primary vote swing in the Reps was mainly larger just because there were more parties running that did not run in 2019.  Rather than being caused by Tudge's personal life, the exaggerated swing seems to have been partly down to Aston's large cohort of Chinese-Australian voters (14% of residents, though a smaller percentage of voters).  

I have seen some comments suggesting United Australia Party preferences were crucial to Tudge's victory in 2022.  This is false.  The UAP did poll a high vote of 6.1%, with a high flow of 69.3% to Tudge, both these figures being bolstered slightly by the donkey vote.  However Tudge's victory margin (5520) easily exceeded his margin on UAP voter preferences (2312).  Had Tudge received no UAP voter preferences at all he would have lost but that was never going to occur.   The rate at which preference recommendations are followed by UAP voters is low so that the deregistered party's expected absence from the by-election is not likely to have much impact.

The evidence suggests that 2022-model Tudge had a pretty standard incumbent personal vote.  His 2PP was a couple of points higher than the Senate 2PP for his party. 

Teal encore?

The media's obsession with, and at times semi-naked spruiking for, independents is such that no sooner had the by-election been called than the teal alarm bell was being sounded on account of Tudge's indifferent primary vote at the last election.  It's interesting that no teal ran against Tudge given how central he was to criticism about Coalition MPs' attitude to women.  But that's history now and there's no special traction for a teal running on those issues now that Tudge has gone.  I really don't expect tealism to be a factor at this election - firstly it's just not that sort of seat (it's not really inner city or a seat with a very high share of highly educated voters) and secondly I wonder whether the protest value in voting for teals is still the same with the Morrison government gone.  Thirdly because Labor is competitive in Aston, it will be harder to run a strategic voting argument for voting teal (indeed the preference flows seen at the federal election weakened that argument anyway).  

[Update: Indeed, no teal contested the seat.]

Dutton Factor!

Expect to see a lot of talk about this one as a test for Peter Dutton, though voters don't like being talked down to and told what they have to think about in a by-election and it may well be about completely different factors.  The theory is that Peter Dutton is particularly unpopular in Victoria and that voters might use the by-election to pressure the Liberals to dispose of him.  There's good support for the theory that Dutton is not liked in Victoria in the recent Newspoll quarterly figures which found him at net -15 in Victoria compared with net -1 in South Australia although the 2PP in both states was the same.  There's no public evidence I've seen on whether Astonian voters specifically share this sort of view or are likely to use the by-election to damage Dutton. However Dutton has enough China-hawk form to suggest some reason for nervousness about whether he will repair the damage among Chinese-Australian voters.  If there's a clear risk there the Liberals will just keep their leader away beyond token appearances and focus on their candidate and local personnel.   

Late in the campaign Labor has rolled out anti-Dutton material, but what I've seen of it seems not especially compelling (including, eg, that he was unsuccessful in leadership bids in 2018 - so what?)

Assessment

This section will be edited as more information emerges but an early assessment is that Aston really should be retained if the Coalition can select a non-disastrous candidate.  It might be retained fairly comfortably or there might be a small swing to Labor.  Yes Labor is polling well, but the fact that it is coming off such a high election base doesn't help in terms of driving a further swing.  The shift in Aston at the last election may have been partly temporary and its recent history as a safe seat probably still counts for something.  

Assuming it comes down to a 2PP contest (and the early signs are Labor is running), my initial benchmarks are:

56+ to Liberal: Good result for Liberals

52-56: Not much to see

sub-52 but clear on the night: Good result for Labor but consistent with their known popularity

Liberals win but not clear on the night: Very good for Labor

Liberals lose: Very bad for Liberals (especially if to Labor) but perhaps not the disaster some are saying - every by-election is different and Labor's honeymoon has to end someday

As of March 30 the Liberals are facing distractions with the Victorian Liberals involved in an internal mess over the disciplining of state Legislative Councillor Moira Deeming.  Deeming appeared at a rally alongside anti-trans extremists who have frequently appeared with far-right company (supposedly always by accident); neo-Nazis were also attracted to the rally.  There is also some perception that the NSW election exposed general damage to the Liberal Party brand and suggested that the issue with Chinese-Australian voters have not yet gone away.  Further, Campbell was accused by Labor of hypocrisy over childcare.  Maybe none of this actually matters but the perception that the Liberals are in a bumpy environment has seen betting odds close up and a bit more interest in a previously boring campaign.

The Liberals for their part are focusing on cost of living issues and inviting voters to protest against infrastructure cuts in the area in the first Labor budget.

There has been some attention on turnout as both prepolls and postals are lagging compared to 2022.  There is a general assumption that low turnout hurts Labor though I have not seen any empirical evidence to confirm it.  Labor is working hard to try to remind voters to vote.  

Candidates

The major party candidates are Roshena Campbell (Liberal) and Mary Doyle (Labor).  Doyle was the candidate for Labor in 2022.  I will mention any other candidates of note or amusement value.    Doyle, a union organiser, is local to the general area but does not live in the electorate.  Campbell (a barrister, Melbourne councillor and Age opinion writer) is an import from the inner suburbs but has announced she will move to the electorate immediately if she wins.  She also claims a connection to the area from defending residents in the past.  Doyle appears to be a safe Labor pick having far from disgraced herself in the 2022 campaign,  but not a particularly high profile name.  Campbell is the more double-edged choice: politically experienced, high-profile, well connected and reportedly a fine speaker, but also from outside the area and whose past commentary will be raked over for controversial views.   

Update 10 March: A small field with the other three candidates being the Greens' Angelica Di Camillo, Fusion's Owen Miller and independent Maya Tesa.  Tesa is a recent candidate for the Liberal Democratic Party but the LDP is deregistered federally for the time being as the endgame to exploiting a loophole to keep its old name on the ballot for 2022.  She is running as an "Independent Libertarian".  Fusion is a merger of various former left micro-parties (Science Party, Secular Party, Pirate Party, Climate Emergency Action Alliance: Vote Planet and Australian Climate Justice).  

The ballot order is Miller, Campbell, Di Camillo, Doyle, Tesa so Campbell gets the advantage of higher ballot position (probably worth a few tenths to half a percent.)  Given the lack of right-wing opposition beyond Tesa, Campbell will probably need a higher primary than the UAP-dented 43 Tudge recorded in the general election.  

On the day before election day there was publicity about Doyle apparently incorrectly filling out a Section 44 declaration form by answering "N/A" to the question "Q7. Do you know which citizenships have been held by each of your current or former spouses and similar partners?"  She also did this on her 2022 declaration, but was formerly married, so the answer should have been either Yes or No.  There is no evidence she is actually ineligible, and if elected she will get a chance to correct the record.  

Polling

No neutral polling has been seen.  On 30 March Sky reported a supposedly "leaked" (translation: given away) Labor internal poll said to show Labor trailing 48-52 with hugely negative personal ratings for Peter Dutton (21-50) and very positive personal ratings for Anthony Albanese (56-26).  While any or all of this may be true, such internal polls only surface when they tell the story the party wants them to tell, and should be treated with great caution (if they exist) for this reason.  As usual no details of pollster, sample size etc and the poll is over two weeks old.