Preference Flows at the Secret Harbour By-Election

Update after full distribution of preferences published: The Liberal Party polled 4,777 votes. During the distribution of preferences, its tally rose to 6,018 votes. When excluded, 3,549 votes transferred to One Nation. That means it is mathematically impossible for Liberal preferences to flow to One Nation at a rate higher than 74.3%. (3549/4777).

My original estimate based on flow data counted out for the Early Voting Centre was that 70.2% of Liberal preferences flowed to One Nation. While I cannot update that number from the distribution of preferences, the lower Liberal vote on polling day suggests to me the % flow of Liberal preferences to One Nation is under 70%.

A quick break from my holiday to deal with some nonsense claims made about preference flows at the Secret Harbour by-election.

In this YouTube video, Topher Field uses some weird maths to come to the conclusion that the Secret Harbour by-election saw a record flow of Liberal and right-wing candidate preference flows to One Nation.

It is a bit hard to be clear what he is saying, but I think he comes to the conclusion that once you exclude left-wing parties from the calculations, 88% of preferences from right parties, including the Liberal Party, flowed to One Nation.

This claim is rubbish and preference flow data released by the WA Electoral Commission shows that claim to be wrong.

The WA Electoral Commission has released the preference tally sheets for the Secret Harbour Early Polling Place, which includes all votes cast in the pre-poll along with votes taken on the day. This comes to 13,049 preference votes, 47% of all votes cast.

The published date is a count out of preferences from the 14 excluded candidates to either Labor or One Nation. The chart below shows the preference flows to the two parties from the five parties polling more than 1%, the other nine candidates accumulated.

The data for this graph can be found on the WA Electoral Commission website linked under the heading “Secret Harbour Early Polling Place Count Results”

(UPDATE: The above data shows an overall flow of preferences from all parties to Labor of 46.4%. On a nearly complete count, the actual flow, calculated by subtracting the primary count for Labor and One Nation from the two-candidate preferred count, produces an electorate total preference flow to Labor of 50.8%. That means of the votes for which preference flow data is not available, 54.2% flow to Labor. While some of this higher rate is due to the Green primary vote being higher on polling day and the Liberal vote lower, it is possible the Liberal flow to One Nation will slip below 70%.)

The flow of Liberal preferences to One Nation was 70.2%. That is very much on par with the preference flows observed at the South Australian election and very similar to the flows seen at Queensland elections where the LNP has recommended preferences to One Nation.

Greens preferences flowed 90% to Labor, a very high figure for an outer suburban seat. The Secret Harbour result confirms that preference flows between parties of the left continue to be stronger than between parties of the right. There is nothing new in terms of preference flows in Secret Harbour.

The real story of the by-election is the primary vote levels for Labor, Liberal and One Nation. The Secret Harbour result is remarkable because One Nation polled so strongly and Labor was very badly defeated. But in terms of preference flows to One Nation and Labor from the Liberal Party and other candidates, there is nothing ground breaking or remarkable. We’ve seen these preference flows before.

As I said above, the preference flows in Secret Harbour were the same as we saw at the South Australian Election and at previous elections where Liberal preferences were distributed.

The flows in Secret Harbour are roughly what pollsters have been using for some time, either based on past results or based on what survey respondents answer.

There is nothing new in the Secret Harbour preference flows, nothing to suggest the pollsters are getting preference flows wrong, just another confirmation that around 70% of Liberal preferences flow to One Nation.

FURTHER COMMENT: A few people were having a go at me on X arguing that Topher Field was disadvantaged by not having access to the preference data.

No. His analysis was wrong because he adopted a series of preference assumptions about Greens and other left-wing parties that everything we know about preferences tells us were wrong.

He assumes Greens and left-wing preferences flow 100% to Labor. That does not and never has happened. As the graph in this post shows, 10% of Greens preferences flowed to One Nation, and even higher rates were recorded for parties that Field classified as left-wing and treated as 100% to Labor.

That means that when he turns his attention to right-wing preferences, he has inflated the flow of preferences to One Nation because the total he starts with INCLUDES the Labor and left-wing preferences to One Nation that he has assumed didn’t exist.

That’s why Field comes up with his 88% figure and a claims a record high flow of right preferences. His figures comes about because he has ignored every available previous example of preference flows, assumed 100% left preferences to Labor, and that wrong assumption has led him to produce a flow of Liberal and other right party preference flows to One Nation way above what actually happened at the by-election.

As I said several times in this post, the Secret Harbour result tells us a lot about the attraction of One Nation in certain types of Labor seats. But the result doesn’t tell us anything we don’t already know about preferences, that the rate of Liberal voters preferencing One Nation is about 70%.

On another note, in his discussion about the Farrer by-election, he uses the wrong data to analyse preferences. Here is the actual preference flow data published by the AEC.

There is a 10% difference between Liberal and National preferences. Some of this will be due to the Independent coming from the Liberal end of the electorate, but it does indicate that there appears to be some reticence from Liberal voters to preferencing One Nation, and there was also some evidence of that in the South Australian election results.

And on one final point that annoys me not so much in Field’s analysis, but in all current analysis of opinion polls. The next election will not boil down to a Labor versus Coalition contest, or a Labor versus One Nation contest. All across the country you will see different combinations of final candidates in different contests. And each individual contest will deliver its own seat winner.

No single after preferences national count can possibly indicate how many seats each party will win which is why I’m not sure why everyone is so obsessed about getting the estimates right.

I think one advantage Labor has is that the Coalition and One Nation will poll most strongly, and deliver each other the strongest preference flows, in seats Labor is unlikely to ever win and will often finish third. These seats are in almost all cases currently held by the Coalition.

There will be seats where One Nation or the Coalition will together challenge Labor. The Secret Harbour result reveals an example. But in seats where Labor leads on primary votes, Labor will be assisted by preference leakage between its two opponents. And I expect the flows to be weaker in these sorts of seats, most of which are in the city, than we will see in rural seats where Labor is less likely to win.

As I outlined in a previous post, I view the coming Victorian election as the most important election in my lifetime for the structure of the Australian party system. It is a system that for so long has been two-party dominant but shows every sign of fracturing.