Category: Federal By-Elections

10.17pm. Thornbury pre-poll also swings slightly to Labor.

10.00pm. The Bundoora pre-poll voting centre has reported, and it produced a result typical for the electorate in swinging slightly to Labor. We will presumably get a further three pre-poll voting centres this evening, and presumably also a batch of postals.

8.36pm. Only pre-poll voting centres now outstanding, on the primary vote at least.

8.27pm. Labor’s small but seemingly decisive lead is holding more or less firm, now at 2.9% on my projection. Most of what remains is the large pre-poll voting centres.

8.17pm. The tide keeps flowing to Labor, with my model (3.4%) now more bullish for them than the ABC’s (1.9%).

8.13pm. Not sure where exactly, but a very good result has come in for Labor, pushing their lead out to a near-insurmountable 2.9% on my projected measure, which now differs only slightly from the raw result of 3.2%.

8.08pm. Most of the polling day booths are in now, and I’ve got Labor’s lead firming very slightly. If the Greens have a hope, it’s that a different dynamic will play out in the pre-poll voting centres.

8.01pm. As the count slowly creeps up, Labor retains its lead of around 1.5%. The Greens will want a couple of good results to come through pretty soon.

7.56pm. With around half the booths now in, the distinction between my model and the ABC’s has all but disappeared: Labor holding in both cases with a 1.5% margin.

7.54pm. Now I’ve got Labor’s lead out to 1.4%, which is a handy place to be with 40% of the vote counted, but not yet bolted down.

7.45pm. Yet another change of lead on my projection, but I’ve got the lead at 0.9% compared to the ABC’s 1.5%. ABC still staying Labor retain, I’m still saying too early to call.

7.42pm. The ABC computer is calling it for Labor, but it’s making no effort as I am to project preference flows on to seats where only the primary vote has reported, which is around half of them.

7.38pm. My hope that this might be sorted early on and I could devote my energies to South Australia is not being realised: once again my projected lead has changed hands, in large part because Labor’s preference share has now declined to 65%.

7.35pm. Now with over a quarter of the vote counted, Labor leads on the raw vote, but I’m projecting that to come back a little. Very close, in a nutshell, but Labor slightly favoured.

7.31pm. And now I’ve got the Greens with their nose in front. There are six booths in on two-party, none of which have swung much, but big swings to Labor in some of the booths in which we only have two-party numbers. The preference flow from the latter is being projected on to the former, and I’ve got Labor getting 69.4% of them.

7.25pm. Now with more substantial numbers in, it’s looking very tight – absolutely no swing at all on my two-party projection, with 12 counted in primary and five on two-party.

7.19pm. Some better results for the Greens send the pendulum back their way. My preference model is now going entirely off results from this election, and Labor is receiving 71.9% of them — 337 to 132 to be precise, going off the three booths that are in on two-party.

7.17pm. Labor have had some thumpingly good results in Alphington North and Collingwood, such that the ABC is projecting a 6% swing in their favour, and I’m projecting 7%.

7.11pm. Been sorting through a lot of technical problems with my live reporting, and I think I’m past the worst of them. So we’ve got three booths in on the primary vote and two in on two-party preferred, and my assessment is that it’s looking tight but with the Greens with their nose in front. For the time being though, this is largely based on a preference flow derived from the Northcote state election result, and the very early indications are that they might do better than that.

7pm. A very small polling booth called Murray, with 249 votes cast, has the Greens up 8.5% on the primary vote and Labor steady.

6pm. Polls have closed for the Batman by-election. This being an inner-city seat with large booths, it should take a while for us to start seeing numbers – perhaps as long as an hour. Wish me luck with my live results reporting facility.

Click here for more detailed (and better formatted) results.

Two days out from the Batman by-election, a dedicated thread for discussion. The campaign has been dominated by the disunity of the local Greens, with Alex Bhathal’s opponents characterised by Noel Towell of The Age as “long-term party stalwarts … aghast at some of the people surrounding the candidate, who they say are newcomers to the party with a ‘whatever-it-takes’ approach to their politics”. The Financial Review reports that the Greens’ infighting has given Labor some hope in what is generally acknowledged to be an uphill struggle, with Ladbrokes presently offering $1.45 on the Greens and $2.50 on Labor. However, former Labor Senator Stephen Conroy offered a pessimistic view on Sky News yesterday.

Two notable developments in the Batman by-election campaign:

• The Age has results of a poll conducted by Lonergan Research which, unusually these days, targeted only landline phones. The poll of 700 respondents found Labor leading 53-47, from primary votes of Labor 40%, Greens 39%, others 16% and don’t know 5%. However, it might be thought the lack of mobile phone polling could skew the result in favour of Labor. It should also be noted polling ahead of the Northcote by-election in November understated support for the Greens. The poll also found 36.2% of respondents had the impression Labor supported the Adani coal mine, 28.3% believed it did not, and 20.3% did not know, with 76% saying they would be more likely to vote for a candidate who opposed the mine, against 24% for less likely.

• With this week’s closure of nominations, it emerged that ten candidates had come forward, which is at the low end of normal for federal by-elections these days. The ballot paper order runs Yvonne Gentle (Rise Up Australia), Ged Kearney (Labor), Alex Bhathal (Greens), Kevin Bailey (Australian Conservatives), Tegan Burns (Australian People’s Party), Debbie Robinson (Australian Liberty Alliance), Teresa van Lieshout (Independent), Adrian Whitehead, Mark McDonald (Sustainable Australia) and Miranda Smith (Animal Justice).

The text is a work in progress, but I have a Batman by-election page up, which will keep score ahead of the by-election initiated by David Feeney’s resignation announcement yesterday. This is replete with the full English of charts and maps detailing past results and the electorate’s demographic characteristics. The highlights for my money are a chart showing the rather extraordinary progress of the primary vote since the Greens first fielded a candidate in 1996, and a booth results map that shows the neat precision with which Bell Street divided the booths the Greens won in 2016 from those won by Labor. If you would like more of my thoughts on the subject of the Batman by-election, there will hopefully be an article on the subject from me in Crikey today.

Projected ALP swing	Projected 2PP	ALP win probability

	PRIMARY VOTE			TWO-PARTY PREFERRED
	#	%	Swing	#	%	Proj.	Swing
Robinson (ALA)	623	0.9%
Folitarik (SPP)	827	1.1%
Jansson (FUT)	902	1.2%
Alexander (LIB)	31,901	44.1%	-5.9%	39,195	54.2%	54.1%	-5.6%
Keneally (ALP)	26,290	36.3%	+7.4%	33,172	45.8%	45.9%	+5.6%
Platter (APEP)	149	0.2%
Alick (GRN)	5,000	6.9%	-2.0%
Ziebell (AAHP)	622	0.9%
Fels (NCP)	116	0.2%
Richa (ACP)	3,251	4.5%
Cao (CDP)	2,299	3.2%	-3.5%
Golding (AUP)	386	0.5%
FORMAL	72,366
Booths reporting on primary vote (out of 41)							38
Booths reporting on two-party preferred (out of 41)							38
Formal votes counted as % of enrolment (106,582)							67.9%

Summary

All I have to add at this stage is the chart below, which seeks to give some insight into how well by-election swings have worked historically as pointers to the next election result – to which the answer is, not very well at all. Featured are all federal by-elections contested by both the Coalition and Labor back to the Whitlam government, with the by-election swing to the government (nearly always negative) recorded on the horizontal axis, and the subsequent election swing (usually negative as well) on the vertical. As such, all by-elections for a given parliamentary term have the same result on the vertical axis. I have also included a line recording the correlation between the two variables, but only for by-elections that were not held in the first nine months of the parliamentary term, which are usually a lot more favourable for the government. However, the predictive power of the underlying equation is very poor (the r-squared result is 0.0655), as it could hardly fail to be, given the government recorded a favourable swing of 7.4% in New England a fortnight ago.

Continue reading “Bennelong by-election live”

A Galaxy poll for the Daily Telegraph has John Alexander clinging on to a 51-49 lead ahead of tomorrow’s Bennelong by-election, after a poll at the beginning of the campaign had it at 50-50. On the primary vote, Alexander is down two to 40% and Kristina Keneally is down one to 38%, with the Greens on 8%, Australian Conservatives on 7% and Christian Democratic Party on 3%. The sample is only 524, but the result is in line with a similar poll conducted by the same company but badged as Newspoll for The Australian earlier in the week.

The Sydney Morning Herald today has results from a ReachTEL poll for the Bennelong by-election, which credits John Alexander with a lead of 53-47 on respondent-allocated preferences – a swing to Labor of nearly 7%. The primary votes, after allocating a forced response follow-up from the (unusually small) 2.4% who initially professed themselves undecided, are 41.3% for John Alexander (down 9.1% on the election), 36.3% for Kristina Keneally (up 7.8%), 7.5% for the Greens (down 1.6%) and 14.9% for the rest. The poll was conducted on Tuesday from a sample of 819. This is the second ReachTEL poll of the campaign, the first being conducted a month ago and showing Alexander leading 54-46. The other two published polls, a Galaxy poll at the start of the campaign and Newspoll this week, both had it at 50-50. Multiple reports suggest party polling has been nearer to ReachTEL’s findings.

For all the background you could want, my Bennelong by-election guide is now updated and much expanded.