Last month, when I reported on the Farrer by-election, I promised on Facebook that there would be “More to come next week after the dust settles.” This post is a belated redemption of that promise.
The actual result barely changed from the initial figures. One Nation’s David Farley won the seat with 57.6% two-candidate-preferred, a margin of about 15,300 votes, off a primary vote of 39.5%. But it was only last week that the electoral commission released the table we’ve been waiting for, showing the flow of preferences that determined that result between him and his opponent, Tealish independent Michelle Milthorpe.
Antony Green and Ben Raue have already blogged about this, but it’s well worth a look given the most controversial aspect of the Farrer campaign: the fact that both Liberal and National parties issued how-to-vote cards recommending that their voters give preferences to One Nation. How much effect did those preferences have?
In aggregate, 58.5% of preferences flowed to One Nation. Out of the ten candidates besides Farley and Milthorpe, preferences from five of them favored Milthorpe, ranging up to an extraordinary 91.8% from the Greens. The other five, all on the right, favored Farley, but the only ones that really mattered were the Coalition, whose preferences flowed 63.5% to Farley: 59.0% from the Liberals and 69.1% from the Nationals.
So National voters, as you might expect, are reasonably comfortable with their votes going to One Nation, although it’s still not an especially tight flow. Liberal voters, rather less so. For comparison, at last year’s federal election, in the nine seats where Liberal preferences were distributed between ALP and Greens they flowed on average 72.1% to the ALP – and they were mostly the sort of seats where you’d think how-to-vote coverage would be pretty poor. So 59% suggests considerable disaffection.
Which is what you’d expect: with the party having turned to the hard right and signalled that it regards One Nation as an ally, there was little reason for voters who agreed with that position not to just vote for One Nation. The remnant Liberal vote (which was less than a third of what it had been last year) was likely to be disproportionately sceptical of One Nation.
There’s a big geographical separation, too (you can download the figures here). In the cities, Albury and Griffith, and in the densely settled irrigation belt near Griffith, roughly half of Liberal voters ignored the how-to-vote card. In the four Leeton polling places, some 56% of them actually preferenced Milthorpe. But in the rural areas along the Murray outside of Albury, One Nation had more appeal, getting more than 60% of Liberal preferences; in the Mulwala, Barooga and Echuca-Moama pre-poll centres that figure topped 70%.
This table that I’ve prepared shows the pattern:
| Liberal preference flow to One Nation | ||
| On the day | Pre-polls | |
| Albury (city & inner suburbs) | 52.8% | 63.1% |
| Lavington (Albury northern suburbs) | 53.6% | 62.8% |
| South-east rural (Corowa, Culcairn, Howlong, etc.) | 64.7% | 63.8% |
| Griffith (urban area) | 48.3% | 51.8% |
| North-east rural (Hay, Leeton, Narranderra, etc.) | 45.8% | 46.8% |
| South-west rural (Deniliquin, Finley, Moama, etc.) | 61.2% | 70.2% |
| North-west outback (Balranald, Hillston, Wentworth, etc.) | 58.3% | 69.3% |
| Postals | n/a | 58.6 |
So, did it make a difference? Farley gained a net 6,049 votes from Coalition preferences; if those votes had instead flowed the other way at the same rate, his margin would have fallen from 15,302 to just 3,204, or 1.6%. Still a win, but now very close. And as noted above, this was not a very tight ticket: if instead we assume a preference flow to Milthorpe more in line with the typical run of Liberal preferences, she most probably would have won.
And that’s without counting the way that the Liberal decision to preference One Nation gave it added respectability, effectively giving permission to unattached voters to look in its direction. Without that boost, it’s fair to assume that One Nation would have struggled.
It’s not entirely clear how much of a lesson we can draw from this for a general election, and not just because by-elections do funny things. Now that the One Nation genie is out of the bottle it won’t be a simple matter to stuff it back in. And assuming the same preference recommendations, it’s likely, as Green says, that “both National and Liberal preference flows to One Nation would have been higher in a final contest involving Labor rather than an Independent.”
That said, the fact that many in the dwindling pool of Coalition voters are reluctant to preference the far right is an important data point. It means that you can’t look at an opinion poll showing, let’s say, One Nation on 30% and the Coalition 20% and just add the two together to get half the vote. The reality is that in a situation where its candidates are being eliminated, a lot of those Coalition votes are going to end up with Labor, Greens or Teals rather than with One Nation.
And perhaps, one day, the knowledge that its voters are not on board with the strategy of appeasement may even deliver a small amount of backbone to the Liberal Party.
The World is Not Enough 