Vasse by-election live

Sunday

The table below summarises the results for the Busselton booths, the non-Busselton booths, and the current total inclusive of postals and pre-polls with swings calculated off the final result from 2013.

Clearly Peter Gordon’s local recognition in Busselton stood him in very good stead, the increase in the Nationals’ vote being 10% higher than elsewhere. I don’t find it quite as easy to explain why the swing to the Greens was substantially larger outside Busselton – certainly not because of homeless Labor voters, there being quite a few more of them inside Busselton than out. The biggest swings to the Greens were in Dunsborough and Yallingup, which are chiefly notable for having a large number of surfers. So perhaps it reflects a hostile reaction to shark culling from those it presumes to protect. The Liberal vote also held up relatively well in these booths, with a smaller swing to the Nationals.

All told, the Liberal vote was entirely in line with what I’d been told during the week the party’s internal polling was saying. However, the Nationals did quite a bit better, and the Greens quite a bit worse. The Greens can and will point to the fact that their vote was up by 8%, but the absence of Labor means their 2013 result is not a realistic baseline. Their vote was below the 21.7% combined Labor and Greens vote from 2013, and they would have hoped for better than that. Of the other candidates, Rebels bikie identity did okay to score 4.4% (853 votes), and can draw some amusement from outpolling Australian Christians on 3.5% (680 votes), who had the advantage of the donkey vote. Teresa van Lieshout’s bikini-wearing wasn’t able to boost her any higher than 1.4% (265 votes).

Election night

9.37pm. Final 2PP for the night: Liberal 10,435 (53.45%), Nationals 9,088 (46.55%). The WAEC has published booth-level 2PP on its media feed but not its website, so Antony Green has those numbers and the WAEC site does not.

9.24pm. That tweet being, as you might have expected, a claim of victory. The pre-poll primary vote have been added to the WAEC site, and the 2PP will apparently show them breaking 56-44 to the Liberals. The final two-party preferred looks like it will be about 53-47.

9.15pm. 4750 pre-polls will shortly be added to the count, and indications are that they’re favourable to the Liberals. From the WA Liberal Party’s Twitter account: “With Vasse early in person votes counted, we will make an announcement shortly about the overall result of this by-election.”

7.58pm. The WAEC have blurted out a single two-party result in one hit, and it has the Liberals on 7855 (52.7%) and the Nationals on 7038 (47.3%). Which is closer than I was expecting at the start of proceedings, but no, as they say, cigar. Preferences split 67.3%-32.7% in favour of the Nationals, so I was right to think Warren-Blackwood wasn’t going to be much of a guide. We can still expect maybe 2000 pre-polls to come in this evening, and further postals over the coming days, but with the Liberal lead now established at 817 votes, the only issue is the size of the margin.

7.44pm. 1335 postals added, and the Liberals get 55.2% of them to 22.0% for the Nationals, which is obviously very good news for the Liberals. Now things start getting difficult in doing projections off the 2013 results, as there won’t be any absent votes and there were only 521 postal votes last time. So the raw primary votes are as good a guide as any, and they have the Liberals on 43.3%, which should be more than enough. But it would still be interesting to see some two-party results.

7.30pm. All booths now in. The last Busselton booth was very much like the earlier ones, so nothing I said in the previous entry seriously needs to be revised. Apparently we’ve still got some pre-polls and postals to come. More to the point, we still don’t have any two-party preferred results, apparently because the WAEC wanted to be sure who the top two candidates were before they proceeded. The preference flow is very much a wild card here, but it would have to be very strong indeed to give the Nationals a shot. They would also have to do well on pre-polls and postals. They’ve given it a good shake though, and defied Colin Barnett’s suggestion that the contest was in fact between the Liberals and the Greens.

7.23pm. Two big booths in from Busselton (quick count, BTW), and it’s now more interesting still. In Busselton itself, Liberal are down 19.4% and the Nationals are up 24.6%. My regionalised projection is 41.7% Liberal, 28.4% Nationals, 19.5% Greens. In neighbouring Warren-Blackwood last year, preferences only favoured the Nationals 54-46. If that’s any guide, the Liberals will still get home with a margin of 6%. However, my intuition is that the Nationals should do better than that.

7.15pm. Things have taken a turn for the interesting again with the first two Busselton results (if you count Vasse Primary School, which I do), together with Cowaramup. The Busselton booths have the Liberals down 18.2% and the Nationals up 23.9%. Based on a projection that gives appropriate weight to Busselton and non-Busselton results, I have the Liberals at 42.5%, the Nationals at 28.0% and the Greens at 19.4%. So the Nationals look set to clear the first hurdle of outpolling the Greens, although they would still have to do extremely well on preferences to be seriously in the hunt.

7.06pm. Yallingup and the two Dunsborough booths are now in, and while there’s still nothing from Busselton, I think we’re past the point that the Liberals are actually in danger. The booth-matched swing against them is about 7%, coming off a 2013 vote of 57%. The Nationals are up 14%, putting them on course for a bit over 21%, and the Greens are up 10%, suggesting they’ll get to about 20%.

6.59pm. Projecting booth swings on to the 2013 totals, I get 44.2% for the Liberals, 23.6% for the Nationals and 23.1% for the Greens — remembering that half the battle for the Nationals is to finish ahead of the Greens. However, it must also be remembered that we don’t have anything in yet from Busselton, and you would not intuitively expect the swings away from the Liberals to be of the size we’ve been getting from the rural booths, notwithstanding the size of Troy Buswell’s personal vote there.

6.57pm. Yallingup supports the trend of the huge swing being rurally specific: here the Nationals are up 11.6% and the Liberals down 6.1%. A great result here for the Greens, up 36.3%. Their overall swing on ordinary polling booths is 15.3% (UPDATE: Belated correction – those were their total votes, not their swings, which were actually 10.3% in Yallingup and 11.3% overall)

6.47pm. At Rosa Brook Hall, the Liberals and Nationals are both up by 9%, whereas the other three booths have the Liberals down by between 15.4% and 26.5%. So we may be seeing a big shift to the Nationals that’s specific to the more agricultural areas of the electorate.

6.45pm. Another two small booths, rural Carbunup and Rosa Brooks Hall just north of Margaret River, have slightly modified the situation, but in the ordinary polling booths so far there’s a 20.5% swing to the Nationals and a 21.3% swing away from the Liberals. If that’s maintained, it will be very close. But Busselton might well behave differently.

6.38pm. Very similar story at the similar booth of Yoorangillup Hall.

6.36pm. The first ordinary booth result is from rural Acton Park Hall, and while it’s only 249 votes, it’s very interesting: the Nationals vote is up from 6.1% to 35.3%, and the Liberals are down from 75.4% to 51.0%. If those swings are in any way reflected in Busselton, we’ll have a contest.

6.24pm. 118 Special Institutions and Hospitals votes are in. These were unusually strong for Labor at the election, so Liberals, Nationals and Greens are all up. However, the Liberals are up most of all, by 14.1%, which would be very encouraging for them.

6pm. Polls have closed, and since a few of the booths are of a small and rural nature, we would should get our first results in reasonably shortly.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics. View all posts by William Bowe