Federal election live: day two

Click here for full federal election results updated live.

Sunday updates

4pm. I gather that in seats where the AEC deemed it had picked the wrong candidates for the TCP count, it has picked new candidates and is throwing to those with the postal vote counts it is conducting today. Presumably it will do new TCP counts for the ordinary votes in these seats in due course, but for now the only TCP figures in these seats are for postals.

That at least is the situation in Bradfield, and I’ve rejigged everything for that seat so the page shows whatever results are available from the fresh TCP count rather than the redundant Liberal-versus-Labor one. The independent contender, Nicolette Boele, hasn’t done nearly as well on the postals (13.8%) as the pre-poll (24.6%) and polling booths (23.1%). She has however received nearly three-quarters of the preferences from the postals, which leave her short by around 52-48 if applied to the overall results. However, she will presumably lose further ground as more postals are added.

I’ll now to through all the other seats where I’d have thought this might be happening and will add commentary as I go.

End of election night

A lot more naturally remains to be said, but as was the case with the commentary I offered here last night, I’ll let my results system do the talking and mostly limit myself to explaining what it’s up to. First, I should note that my reading of the national two-party preferred – 52.8-47.2 in favour of Labor – is 1% stronger for Labor than the ABC’s, which is quite a lot closer to Nine’s projection, and I’m not going to pretend you should take my word for it over theirs. I’ll have a closer look at this tomorrow, but given the state of the primary votes and the large number of votes still outstanding, a lot of this is necessarily based on projection.

As you can see at the top of the entry page, I’m currently calling 71 seats for Labor, 46 for the Coalition, two for the Greens, three for independents and one for Katter’s Australian Party. However, this largely reflects the fact that the system is very cautious in making calls in most of the seats where independents and the Greens are looking formidable, for reasons that will be explained shortly.

Bennelong, Gilmore, Deakin, Menzies, Bass, Lyons, Sturt, Lingiari, Moore and — to stretch the elastic a little — Monash, Casey and Dickson are the old-fashioned kind of in doubt, with close races between Labor and Coalition candidates that could go either way on late counting. Projections here are based on comparison of how the votes in so far matched up with the equivalent votes last time, and an assumption that postals, absents and out-of-division pre-polls will record broadly similar swings (there are also a fair few pre-poll voting centres around the place that didn’t complete their counting last night, despite the head start they had for the first time in sorting their envelopes from 4pm).

However, it’s possible that postal votes in particular will behave quite a bit differently this time in swing terms due to a roughly 70% increase in the application rate. If that results in the postal voter pool being more demographically representative, these votes could be less conservative than usual. For the time being though, this is only a hypothesis.

Most of the seats that are listed as in doubt are ones where there is no historic data by which this result can readily be compared with the last, as is the case in the many seats where non-incumbent independents are in the race. My system applies a wider margin of error in these cases, meaning a fairly substantial lead is required before the 99% probability threshold is crossed and the seat is deemed to be called. For example, no human observer doubts that Allegra Spender has defeated Dave Sharma in Wentworth, but my system isn’t all the way there yet.

The AEC has pulled its two-candidate preferred counts in 15 seats where it deems it picked the wrong candidates, although it seems this hasn’t caused the results to disappear from my pages, which I believe is a happy accident. My two-party projections in these seats work off my estimates of how preferences will flow to the candidates deemed likely to finish first and second. There are a few here that warrant explanation or discussion:

Brisbane. While it is clear LNP member Trevor Evans will lose this seat, it is not clear to whom he will lose out of the Labor and the Greens candidates. The AEC conducted a traditional two-candidate count between the major party candidates, but has discontinued it because Labor is running third. However, it’s far from clear that it won’t prove to have had it right the first time after all the votes are in. There would not be much point in a fresh two-candidate count here as it is already clear who would win between the Greens and the LNP. It would be more instructive to determine how many preferences from the minor candidates are going to Labor and how many to the Greens. Most likely we will not be sure of the result here until the full preference distribution is conducted, which can’t happen until all the votes are in.

Griffith. Here too Labor is running third, but in this case it’s the Greens first and LNP second rather than vice-versa. Labor’s Terri Butler could theoretically make it over the line on preferences if she moved in to second, but the flow of One Nation and United Australia Party preferences to the LNP are likely to put paid to that.

Ryan. My projection of a 52.6% Greens two-party vote over the LNP is based on an estimated 85-15 split of Labor preferences, which the ABC seemingly expects to be even wider, because it has it at 53.7-46.3.

Macnamara. Labor, the Greens and the Liberals are very close on the primary vote, and to the extent that the Liberals are slightly behind, I’m projecting them to make up most of the gap on late counting. So any three of them might end up being excluded before the final count. I would have thought Labor would win on either scenario where it clears this hurdle, since Liberal preferences favour them over the Greens fairly solidly when they direct their preferences that way. However, the ABC is projecting a lineball Labor-Greens result, so perhaps I’m wrong. If so, a fresh two-candidate count between Labor and the Greens would be instructive, but it would only apply if the Liberals did indeed go out before either of them. Certainly the Greens will win if it’s Labor that gets excluded.

Cowper. Independent Caz Heise landed well clear of Labor, and with Nationals member Pat Conaghan well short of a majority at 40.4%, my projection is that she will take it right up to him after preferences. We will need a two-candidate count to see exactly how accurate that is, which presumably the AEC will be forthcoming with fairly shortly – perhaps as soon as today.

Bradfield. A similar story here, with yet another independent, Nicolette Boele, outpolling Labor to finish a clear second, while Liberal member Paul Fletcher is on less of the primary vote than he would like. My preference estimates suggest Boele won’t quite get there, but here too we will need a new two-candidate count to see if I am right.

Now for a very quick look at the Senate before I collapse altogether. It seems to me there is a fairly strong possibility of what Labor would regard as a rather happy result where they and the Greens between them have half the numbers, and can get the extra votes needed to pass legislation from two Jacqui Lambie Network Senators or ACT independent David Pocock. Each of the six states seem to be looking at results where the first five seats have gone two Labor, two Coalition and one Greens, with the last seat up for grabs. To deal with the latter situation in turn:

New South Wales. Most likely a third seat for the Coalition in New South Wales, unless right-wing preferences lock in strongly behind One Nation.

Victoria. Probably the United Australia Party unless preferences flow strongly to Legalise Cannabis, a possibility I hope to be able to shed more light on after running an analysis on past ballot paper data.

Queensland. Probably Pauline Hanson but possibly Legalise Cannabis, who have been something of a surprise packet across all Senate races but particularly here at 6.7%. Neither Clive Palmer nor Campbell Newman are in contention.

Western Australia. The strongest possibility would seem to be a third seat for Labor, something it has never managed before in Western Australia and indeed hasn’t managed anywhere at a half-Senate election since 2010. However, I will also investigate the possibility that One Nation and Legalise Cannabis might be in contention instead.

South Australia. Very likely One Nation’s Jennifer Game, whose daughter Sarah Game won the party’s seat in the state Legislative Council in March, and has quickly emerged as something of a surprise packet. Out of contention is Nick Xenophon, whose “Group O” managed only 2.7%.

Tasmania. A second seat for the Jacqui Lambie Network, whose Tammy Terrell is on 8.1%, almost matching the 8.3% the party managed with Lambie at the top of the ticket in 2019.

Then there’s the Australian Capital Territory, where in yet another triumph for teal independence, David Pocock seems assured of unseating conservative Liberal Zed Seselja. The ABC projection says otherwise, but it seems to me that Seselja will assuredly be buried by the flow of preferences to Pocock from the Greens and Kim Rubenstein.

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