Thursday, May 23, 2019

2019 Federal Election Postcount: Mallee

MALLEE (Nat vs ALP - 19.8%)

Webster (Nat) has won after the seat remained a Nats vs ALP seat by 386 votes.  It is unknown and will perhaps never be known what would have been the Nat-Lib result had Labor been eliminated in third.  (I expect Webster would still have won, but am awaiting the preference distribution.)

In the election leadup I had my eyes on the Victorian seat of Mallee as the most likely to deliver an absolute mess in the postcount.  On election night it seemed to be a bit of a fizzer because none of the independents made 10% in their own right, making it clear that the Coalition was headed for victory.  However the Mallee count has thrown up some interesting complications, and there is a theory doing the rounds that the Liberal candidate Serge Petrovich might be able to defeat Webster if he can make the final two.  I am unconvinced about this theory, firstly because I'm doubting he will make the final two, and secondly because even if he does a rather strong preference flow is needed to get him over the line. I don't think that will happen, Labor HTV card notwithstanding, but in the meantime there's a possibility Mallee will create electoral history.  Never (thanks to Malcolm Baalman for this) has a candidate who finished fourth or worse on primaries in a federal seat reached the final two, and it is possible that this could yet happen.

Mallee is a rural/regional seat in the north-west of Victoria that includes the cities of Mildura and Horsham.  Four MPs, all of them Nationals, have held the seat since its creation in 1949.  The most recent was two-term MP Andrew Broad, who was brought down by a "sugar daddy" scandal in 2018 and did not recontest the seat.  Already reeling from the tabloid misadventures of Barnaby Joyce, the Nationals sensibly preselected Dr Anne Webster, a Mildura-based sociologist.

Following the success of Voices for Indi and the destruction of the Nationals' rural vote in the NSW state election, there was a lot of interest in independent campaigns for Mallee, the two most prominent of which have been by Mildura deputy mayor Jason Modica and former Yarriambiack (I had to copy and paste that) Shire mayor Ray Kingston.  However indies on either side of the Murray did not fare as well as the Shooters had in NSW, despite strong reports of a feel on the ground for change in Farrer.  In Mallee the problem was always that there was no obvious single indie contender who could concentrate the vote.

Mallee attracted the largest field in the nation, 13 candidates.  This is how the primaries presently stand (click for larger slightly clearer version):

It's a long way to 50% but somebody's got to get there!  We also know that the two-party preferred is currently 66.73% to Webster, meaning that she is currently getting 67.7% of preferences ahead of Labor.  Probably she is getting nearly all the Liberal preferences and about 55% of the rest.

Obviously Labor do not win if the final two are Nationals vs Labor.  Depending on the strength of flow between the independents, and to the independents over Labor from the 15% of minor party votes below them, it's possible either Modica or Kingston (I understand more likely Modica) can get over Labor into third.  However the flows between the indies won't be anywhere near 100% so it's hard to be sure that this will happen.  If one of the indies can get into third, Labor preferences should easily push them into second ahead of Petrovich.

If an indie does make the final two, let's be conservative with the Petrovich flow and say only 85% of it goes to Coalition partner Webster ahead of the indie.  In this case, Webster would only need 13% of preferences to beat the indie and would obviously get that easily (especially with UAP and Shooters in the mix).  Even if there is a remarkably weak flow from Petrovich, eg 75%, then the 83% of the rest the indie would need to win is still not realistic.  So I cannot see how the indies could win; their primary vote is simply too low.

The most interesting final two would be a within-Coalition contest between Webster and Petrovich.  A within-Coalition final two happened in 2013 but in that case the Coalition partners had 66% of the primary vote, rather than 47.3% ! Labor has preferenced the Liberals ahead of the Nationals in Mallee, but Labor voters are not very obedient when it comes to being told to preference the Liberals.  In 1993 according to the ABC over 40% of Labor voters ignored such a card and preferenced the Nationals, and I can find no three-cornered contest recently where Labor preferences have broken to the Liberals even to the extent of 60-40.  If Petrovich doesn't get the preferences from Labor I greatly doubt he'll get them from the rest, especially as Webster is a local while Petrovich isn't.  Thus I am sceptical that Petrovich will gain more than a few percent (if even that) off the Labor votes, but I'm open to being convinced otherwise if anyone has very rigorous scrutineering data.

All I've heard regarding scrutineering so far is comments that preferences are going all over the place irrespective of cards, and that is not a surprising thing in a contest where various candidates have their own home town and their own team of people they know, and where voter understanding of the region's politics would be high.  I'm grateful to Michael DiFabrizio (Sunraysia Daily) for the remarkable information that the top six candidates in the count have all topped at least one of the division's 100 polling booths on primaries!  (See graphical representation in his article.  Is this an Australian record?) Scrutineering this one would be very challenging because of the question about exclusion order and what I would expect to be wild variations in flow across different booths.

As for the future direction, I'm not sure we'll get anything significant before the throw of preferences which I expect could happen the week after next.  In theory the AEC could realign the count, but they couldn't be sure that the final two actually won't be National vs Labor, and worse still, they run the risk that if they do realign it, they might get the wrong opponent again.

Mildura looks more like a curiosity than a contest, but I thought I should cover it in detail anyway.  Barring further major developments, I won't update this post much until the preference distribution is finished.

Tuesday 28 May: There has been a fair amount of postcount drift in Mallee, with Webster's primary now dropping to 28.25% and Petrovich now up to 18.74%.  So Webster's primary lead is down by 1.4 points.

Wednesday 29 May: Also see Ben Raue's maps of the vote distribution.

Tuesday 4 June: Preference distribution should be underway by now as the primary count is finished.  According to Michael DiFabrizio this could take several days.

Tuesday 11 June: News that both Modica and Kingston have been eliminated.  It is not yet confirmed that Labor are eliminated in 3rd.

Wednesday 12 June: Mallee is now the only seat left in the country that has not been declared or scheduled for declaration; the distribution is proceeding slowly.  Still no news today on who has made the final two.

Thursday 13 June: News from candidate Ray Kingston that Labor has finished second by 386 votes meaning Webster has won the seat.  Not sure we will ever know what the National-vs-Liberal margin would be.  We do however have a three-candidate margin from Antony Green:

Webster 39267
Hart 27472
Petrovich 27086

This means that had Petrovich been second, he would have needed about 72% of the votes with Hart (nearly half of which were not Labor's to begin with) to beat Webster.  In my view that's extremely unlikely, HTV card or no HTV card.

There's also a suggestion that more Liberal preferences may have leaked to Labor than is normal in a three-cornered contest.  About 15% of the votes with Petrovich would have flowed to Hart ahead of Webster.  Maybe it was just votes from other sources that were to blame for this.  The usual split from 1 Liberal votes is about 8%. 

Also, Modica was very narrowly eliminated in fourth, by around 200 votes. 


  1. My limited insight into Mallee comes from driving the length of it a couple of weeks before the election...

    A common problem for independents in regional federal seats is that it's hard to get a profile in the whole electorate - at state level local government has often been used as a springboard for successful independents, but being a Mildura councillor isn't going to give you much profile in Horsham or Maryborough. Mallee's an interesting case because you have two independents with bases at opposite ends of the electorate and preferencing each other, which counteracts that problem to some extent.

    Agree that on current numbers Webster should be able to collect enough preferences to win. In some past contests between Libs/Nats and a viable independent in a country seat (e.g. Indi 2013), Labor preferences have flowed at 90+%, but even if that happens here against an opponent with less baggage than Sophie Mirabella, Webster should still get enough preferences from other minors to get over the line.

    It's a little surprising that the Liberals didn't contest the seat more seriously, given they've got close the last couple of times it's been an open seat (unless doing so was part of a deal which involved the Nationals backing off somewhere else?). As far as the destination of Labor preferences in a Lib-Nat contest is concerned, it's worth remembering that in a seat like Mallee a lot of voters will never see a Labor HTV card as they would probably lack the capacity to staff many of the booths.

  2. Very interesting stats. Extremely regional strongholds - Modica could just about win a seat just based on Mildura, Kingston could win a seat if it were just the southern half. But neither able to beat the Nats across the whole of the electorate.

    Need a "Voices for Indi" or Warringah-style operation to pick an independent and run a coordinated effort across the whole seat.