Why are One Nation’s preference flows falling?

People are making a huge fuss about the falling preference flows from minor parties, Liberals, and Labor to One Nation. I’m frankly sick of the convoluted discussion surrounding this topic, somehow implying that traditionally-voting major party supporters are becoming increasingly alienated by ONP policy and rhetoric – is it not obvious that the preference flow changes are merely a statistical artefact?

It’s no mystery that this pool of remaining Liberal voters – the ones who have remained steadfast in their support despite the strong ONP campaigning and messaging – are less likely to support ONP to the same degree as the prior Liberal voting pool, which used to contain these voters now supporting ONP. This is, of course, one reason among many, but its certainly the most significant reason in my view.

While Antony Green’s analysis that “there appears to be some reticence from Liberal voters to preferencing One Nation” is of course true, its really worth noting that there is reticence from remaining Liberal voters, which is a necessary distinction.

Figure 1 demonstrates how preference flows to ONP at recent elections have declined as support for Pauline Hanson increases. This, of course, runs counter to the narrative that these changes to ONP preference flows are the result of declining support for the party among the moderate voting base.

This can be broken down further into where these preferences are falling. I hypothesise based on Figure 2 (below) that, among locations that have a strong ONP first preference performance compared to what would be expected for that electorate, preference flows will be weaker, and the inverse seems to be true as well. The result below is not a statistically significant one, but it is still interesting nonetheless.

It’s not a complicated concept: In short, where people like the ONP candidate, they’ll put them first on their primary vote. If enough people do that, then the remaining ALP/LNP voters are simply the voters that are not inclined to support One Nation. So, where the candidate is better, preference flows will be weaker. I imagine this trend would be interesting to analyse for other parties as well.

Accordingly, the preference flows to ONP at the upcoming election can be expected to be weaker than historical results, but I expect that these flows can be approximated from South Australian results wherever relevant, and conclusions about party-wide support absolutely shouldn’t be drawn from individual electorate preference data, that would be silly.