For the 2022 Victorian election, the Liberal Party has changed its position on whether Labor or the Greens should be listed first on the party’s how-to-vote (HTV) material.
Until the August 2010 Federal election, the traditional Liberal decision was to list the Greens ahead of Labor. At the 2010 Federal election, it was the Liberal preference recommendation that elected Greens’ candidate Adam Bandt as the new member Melbourne.
Some in the Liberal Party were unhappy that Liberal preferences were electing members of a party that Liberals labelled as more radical left than Labor.
The position was changed shortly afterwards and it was announced late in the 2010 Victorian state election campaign that Labor would be listed before the Greens on Liberals HTVs. That has been the party’s position in Victoria at state and Federal elections since.
In most seats the choice is entirely symbolic as Liberal preferences will not be distributed. But there are several inner-Melbourne seats where the Liberal candidate is traditionally excluded during the distribution of preferences. The Liberal decision could have an impact on several Labor-Green contests.
In this post I look at the record of preference flows before and after the Liberal switch on preferences. The question for 2022 is whether the new Liberal position will flip the flow of preference as dramatically as the 2010 decision. Will the decision change the result in any seats?
Federal Election flows
The Australian Electoral Commission (AEC) provides the most accurate data of preference transfers. The AEC tallies preference flow data, that is the total destination preferences flows of ballot papers for excluded candidates. Preference flow data ignores intermediary transfers to other candidates during the formal distribution of preferences. Preference flow data gives you precise numbers on the total votes for each candidate, and total votes transferred as preferences to each of the two final candidates in the count.
Since 2007 there have been 13 instances of Victorian federal seats where the Liberal candidate was excluded in a Labor versus Greens contest. The chart below shows the preferences flow data for these contests.
The change of preference policy occurred after the 2010 Federal election so the first three entries in the above graph were produced by Liberal HTVs recommending preferences for the Greens, and the other 10 entries for Liberal HTVs recommending preferences for Labor.
The low 51.5% flow in Wills at the 2019 election followed the disendorsement of the Liberal candidate. There was either no HTV or the candidate produced his own version.
The average flow of preferences when the Greens were listed first was 18.3% to Labor and 81.7% to the Greens (3 instances) as against 65.9% to Labor and 35.1% to the Greens (10 instances) when Labor was listed first. (Note these are the averages of the flows in each seat, not totals from the seats.)
Clearly the Liberal HTV, or just as importantly the Liberal message on who to preference first, had an enormous impact after 2010. Will the Liberal Party change in 2022 have as dramatic an impact on flipping preference flows back towards the Greens?
It can be argued that the flow against Labor at the Victorian election will be strong based on the Liberal Party's message that voters should "Put Labor Last". Given the Liberal Party may not be handing out a lot of HTVs in inner-Melbourne Left bastions, it is the overall put Labor last message that will have greater impact than HTVs.
That is if Liberal voters accept the new message to vote strategically rather than ideologically.
After a decade of demonising the Greens as too left-wing, it may be that many Liberal voters will still preference Labor first. I'm not sure we will see the 80% flows to the Greens we saw in 2007 and 2010, but anything above 65% would spell danger for Labor.
Victorian Election Flows
The Victorian Electoral Commission (VEC) does not collect preference flow data. The only way to analyse preference transfers in Labor-Green contests is to look at the final exclusion of Liberal candidates during the distribution of preferences. This data includes intermediary preference transfers but is still a reasonable measure of Liberal preference flows before and after the 2010 HTV change.
There have been 14 Labor-Green contests since 2006 where the Liberal candidate was the last excluded, three at the 2006 election where the HTV recommendation was to the Greens first, and 11 since 2010 where Labor was listed first. The contests are shown in the chart below.
Noting first that the data is not as good as the Federal flow data, the same trend still emerges. At the three contests where the Greens were listed first on HTVs, the transfers to Labor were 24.8% and 75.2% to the Greens. In the 11 contests where Labor appeared first on HTVs, the average flow to Labor was 64.9%, to the Greens 35.1%.
It may be significant that the transfers to Labor were much weaker at the 2018 election, though it must be noted that the Liberal vote was down on previous elections. The Liberal Party did not nominate candidates in the four inner-Melbourne seats until the last minute and did not contest Richmond. It is possible the 2018 flows may also reflect antipathy to the Andrews government, which would suggest that actively arguing to "Put Labor Last" in 2022 could have a significant impact.
Will it Make a Difference?
With the Liberal vote being so low in inner-Melbourne seats, a change in Liberal preferences may be less important than the preference flow data suggests.
Below I have modelled the four inner-city seats that for two decades have been Labor-Green contests. I have had to use the old electoral margins for each seat because I am using actual preference counts from the 2018 election to model changes in preference flows. The four seats shown were not significantly altered by the redistribution. In the case of Richmond, I have used the final exclusion of an Independent Liberal as my starting point.
The two scenarios I am using are -
- A reversal of the average flow to Labor since 2010, which would mean a 65% flow to the Greens
- A 75% flow to the Greens, the figure produced by the 2006 election data.
As the results below show, the change in Liberal preferences is enough to flip Northcote from Labor to the Greens using both 65% and 75% flow values.
Brunswick
- Current margin - Greens 0.6% v Labor
- 65% preference flow - new margin Greens 3.2% v Labor
- 75% preference flow - new margin Greens 4.5% v Labor
Melbourne
- Current margin - new margin Greens 1.3% v Labor
- 65% preference flow - new margin Greens 5.0% v Labor
- 75% preference flow - new margin Greens 6.9% v Labor
Northcote
- Current margin - new margin Labor 1.7% v Greens
- 65% preference flow - new margin Greens 1.4% v Labor
- 75% preference flow - new margin Greens 2.6% v Labor
Richmond
- Current margin - new margin Labor 5.5% v Greens
- 65% preference flow - new margin Labor 2.8% v Greens
- 75% preference flow - new margin Labor 1.8% v Greens
As an economist would say at this point, "ceterus paribus", that is I am assuming nothing changes except the Liberal preference flow. I am assuming no change in first preference votes. If the Liberal vote rises, weight of numbers makes the party's preferences more important. If Labor's vote rises in Northcote, then the chances of Labor losing on preferences is diminished. A rise in Green primary vote also makes Liberal preferences less important.
Other seats where Liberal preferences may come into play in 2022 are Albert Park, Footscray, Pascoe Vale and Preston. I have not modelled these seats as there are no reliable preference distributions from 2018, plus none of these were close based on 2018 results.
However, that doesn't mean they won't be close in 2022. If they are, then the Liberal switch of preferences could deliver more than just Northcote to the Greens.
What was the flow in the most recent QLD state election in South Brisbane where they did this?
COMMENT: LNP preferences flowed 64.0% to the Greens at the final exclusion of the distribution of preferences.
I would add Footscray to the list of seats where Liberal preference come into play, given the Greens probably came second in the 3-candidate preferred in 2018 and the redivision has been favourable to the Greens in Footscray (at the expense of the Greens in Williamstown). Still probably an ALP win but on a much smaller margin.
Kevin Bonham has done some calculations estimating that the combination of the redivision and the Liberal`s new preferencing policy may have made Pascoe Vale marginal:
https://kevinbonham.blogspot.com/2022/11/modelling-seat-of-pascoe-vale.html
COMMENT: I mentioned both Footscray and Pascoe Vale in the post, but as an appropriate preference distribution was not available for either seat, I did not include them in my calculation table.
Is the strategy change a calculation that it forces the ALP to spend resources in these seats rather than in ALP-Liberal marginals? Interesting, since the Greens are showing signs that in seats like Ryan in Queensland that they can win seats directly from the Liberals too.
How do you think ‘soft’ swinging voters would respond to the new HTV that may have voted Labor or Greens in the past compared to ‘lifelong’ traditional liberal voters?
COMMENT: I don’t know. I’m happy to see the real numbers on election night.
The latest Resolve Strategic poll gives Labour and Green only 46.3% of the primary vote, but Labour 52.7% of the TPP. I am assuming that 4% of that, around two thirds, will come from Animal Justice, Victorian Socialist and Reason, the bulk of which will flow through the Greens first. If you add to this the potential flow through of Liberal votes then Northcote, Richmond and Pasco Vale should become very interesting races for the Greens.
To my observation handing out HTVs, I do not think that the availability of a HTV matters so much anymore as I have observed (in recent times) scrutineering what I thought was amazingly strict compliance with HTVs preferences on safe booths when my party had no physical competition at the gate. Voters just google the preferences on their mobiles.
I had been meaning to do this analysis for years. Love you Antony! You’ve done it for us!
It would seem that Liberal preferences have not been decisive in delivering a single seat to the Greens. In the three seats won by the Greens with Liberals running 3rd, the Greens led on first preferences; in other seats the ALP held on.
COMMENT: In the upper house Coalition preferences have been important in electing One Nation in Northern Victoria Region. This hasn’t had many mentions.