| ALP | NAT | LIB | GRN | ONP | BROCK | COUNT | |
| PRIMARY | 5041 | 1267 | 7576 | 734 | 134 | 4557 | 19309 |
| % | 26.1% | 6.6% | 39.2% | 3.8% | 0.7% | 23.6% | 100.0% |
| Swing | -16.4% | -7.6% | 0.7% | ||||
| PORT PIRIE | 2157 | 181 | 1344 | 129 | 29 | 2480 | 6320 |
| % | 34.1% | 2.9% | 21.3% | 2.0% | 0.5% | 39.2% | 100.0% |
| Swing | -23.5% | -11.7% | -0.8% | ||||
| REMAINDER | 1368 | 735 | 4292 | 478 | 70 | 757 | 7700 |
| % | 17.8% | 9.5% | 55.7% | 6.2% | 0.9% | 9.8% | 100.0% |
| Swing | -8.6% | -6.2% | 1.9% | ||||
| DECLARATION | 1516 | 351 | 1940 | 127 | 35 | 1320 | 5289 |
| % | 29.2% | 6.5% | 35.7% | 2.3% | 0.7% | 25.5% | 100.0% |
| Swing | -9.2% | -13.9% | -1.9% | ||||
| 3CP | 5532 | 8215 | 5562 | 19309 | |||
| 28.6% | 42.5% | 28.8% | |||||
| 2CP (FINAL) | 9322 | 9987 | 19309 | ||||
| 48.3% | 51.7% |
Thursday, January 28
Malcolm Mackerras muses on this and other recent by-elections in the Canberra Times.
Wednesday, January 27
Electoral commissioner Kay Mousley has officially rejected the Liberals’ request for a recount, on the basis that specific concerns about the counting of votes had not been identified. The mere closeness of the result was deemed insufficient grounds for a recount. Below is the piece I wrote for yesterday’s edition of Crikey, previously available to subscribers only. Martin Hamilton-Smith’s office has been in touch to dispute the claim that the “super Saturday” concept referred to below was seriously considered, saying it came down to “one MP” who had been “canvassing the notion to media”.
For psephologists and related species of political tragic, by-elections can’t happen often enough. But for normal people, forced mid-term visits to the polling booth rank somewhere around brain surgery on lists of favourite things. No political operative should ever need reminding of this, but it appears the South Australian Liberal Party did – and now has been, in terms it won’t forget in a hurry.
Saturday’s preference count for the Frome by-election, held a week earlier upon the retirement of former Premier Rob Kerin, gave the Liberal Party the rudest of shocks three days after it had issued a press release claiming victory. Both Liberal and Labor scrutineers were convinced that Liberal candidate Terry Boylan had survived an early scare, thanks to Nationals voters who ignored the party’s recommendation to direct second preferences to independent candidate Geoff Brock. It was believed this would prevent Brock from getting ahead of Labor’s John Rohde, resulting in his exclusion at the second last count. That being so, the State Electoral Office’s indicative two-party count pointed to an unconvincing final Liberal margin over Labor of 1.7 per cent.
However, it seems scrutineers obsessing over the Nationals had neglected to consider the actions of Greens voters, who in the absence of guidance from the party’s how-to-vote card were thought to have followed their normal practice of putting Labor second. In fact, 42 per cent of Greens preferences flowed to Brock against 37 per cent to Labor – enough for Brock to emerge a bare 30 votes ahead of Rohde, before storming home on Labor preferences to defeat Boylan 9987 votes to 9322.
Before the evening was through, a Liberal Party that could previously be heard expressing nothing but warm goodwill about their good mate Kero suddenly found voice to complain about the “obscure” reasons given for his retirement, which had “fuelled resentment” among voters. However, this was clearly wisdom after the event.
Last June, The Advertiser’s Greg Kelton reported that “senior Liberals” were “hatching a plan which would force the Rann Government to face a ‘super Saturday’ of by-elections on the growing political row over changes to country health services”. This would involve the simultaneous retirement of Kerin (who was quoted saying the idea had been “mentioned a few times”) along with fellow Liberal veterans Graham Gunn and Liz Penfold, initiating by-elections in the country and outback seats of Frome, Stuart and Flinders. As bad as Frome has been for the Liberals, it appears that only the reluctance of Gunn and Penfold to bring forward their retirements has spared them a self-inflicted triple-barrelled disaster.
For all that, Labor shouldn’t get too cocky (and reports from The Advertiser that “gleeful Labor MPs have run off copies of Mr Hamilton-Smith’s ‘Liberal victory’ press release to hold up when State Parliament resumes next month to goad the Liberals” do not bode well in this regard). The two-party swing Labor would have picked up if Brock had run third had less to do with voters’ conscious preferences than with their adherence to how-to-vote cards, which in Brock’s case had Labor third and Liberal fourth. The 16.4 per cent of voters who deserted Labor might very easily find less benign ways to register their evident displeasure with the government when the next election is held in March 2010.
Labor MPs would do well to acquaint themselves with a forgotten episode of Western Australia’s recent political history known as the Peel by-election, which in February 2007 gave Labor a morale-boosting 1.0 per cent two-party swing from a strong performance on the primary vote – for all the good that did Alan Carpenter 18 months later.
Tuesday, January 27
Crikey subscribers can read my by-election post-mortem here.
Sunday, January 25
Electoral commissioner Kaye Mousley refuses a recount. Mousley argues that “the final difference between the two candidates is some 600 votes with the distribution of preferences”, although the point surely is that Brock survived the second last exclusion by 30. That would leave the Court of Disputed Returns as their only recourse. However, the Electoral Act empowers the court only to anoint a different winner or order a new election, and I’m not aware of any basis on which such an order could be made.
Saturday, January 24
7.15pm. The last trickle of 265 postal votes had little bearing on the result: 147 (55.5 per cent) went to the Liberals, 47 (17.7 per cent) to Labor, 37 (14.0 per cent) to Brock, 23 (8.7 per cent) to the Nationals, 10 (3.8 per cent) to the Greens and 1 (0.4 per cent) to One Nation. In other words, they added 10 votes to the hurdle faced by Brock to overtake Labor. Meanwhile, the Poll Bludger has maintained its dismal record in predicting by-election results with this clanger from January 9: “Despite a preference swap between independent Port Pirie mayor Geoff Brock and Nationals candidate Neville Watson, there seems little reason not to think Terry Boylan will easily retain the seat for the Liberals.” That said, there’s plenty of humble pie to go round.
6.55pm. The Advertiser now has a full report, which tells us “Liberal officials say they will be ‘seeking clarity’ on the count from the State Electoral Office”. Also:
Liberal MP for Morphett Duncan McFetridge partly blamed Mr Kerin for the loss, saying he had given obscure reasons for leaving politics which fuelled resentment by voters towards the by-election.
True enough, but I hadn’t heard anyone in the Liberal Party complain before. Indeed, it seems they were happy to bring on the by-election because they were expecting Labor to suffer a bloody nose over the country health plan. In June we were hearing this idiotic talk emanating from the Liberal camp (courtesy of Greg Kelton of The Advertiser):
SENIOR Liberals are hatching a plan which would force the Rann Government to face a “super Saturday” of by-elections on the growing political row over changes to country health services … The move would involve three Liberal MPs in rural seats – who are all due to retire at the next election – stepping down to force by-elections. The MPs, Rob Kerin in Frome, Liz Penfold (Flinders) and Graham Gunn (Stuart), have all been outspoken in their criticism of the Government’s planned changes to rural health services … Mr Kerin told The Advertiser the by-election idea had been “mentioned a few times’” but he had not spoken to anyone about stepping down in Frome which he holds with a 4.2 per cent margin. He said he would not rule out the idea … (Gunn) ruled out stepping down to force a by-election in his seat of Stuart which, with a 0.4 per cent margin, is the most marginal Liberal seat in the state. Ms Penfold, whose vast Eyre Peninsula seat of Flinders is the safest Liberal seat in the state, said normally she would not support any moves for a by-election. “But this is such an important issue I will reserve my judgment,” she said.
6.45pm. The surprise packet was the flow of Greens preferences to Brock – 41.7 per cent against 36.6 per cent for Labor and 13.4 per cent for the Liberals. The estimates I was using in my preference calculation were 30 per cent, 50 per cent and 20 per cent respectively. The reason Brock was being written off was the high number of Nationals voters who were defying the HTV card and preferencing Boylan. The Nationals preference distribution I eventually arrived at based on Antony’s reports of what scrutineers were saying was pretty much accurate: 48.0 per cent to Brock (I had 45 per cent), 37.8 per cent to Boylan (I had 40 per cent, which admittedly was the low end of what Antony was expecting) and 14.1 per cent to Rohde (I had 15 per cent). No doubt the page on the Liberal website on Wednesday claiming victory will be removed shortly, so I’ve preserved it for posterity here. That said, we may yet get a recount.
6.20pm. Wasn’t looking hard enough – SEO preference distribution here. The amazement lies in the second last exclusion: Boylan 8215, Brock 5562, Rohde 5532. With Rohde excluded, preferences give Brock his 1.7 per cent victory.
6pm. BROCK SHOCK! Nothing yet on the SEO or Antony Green’s site, but The Advertiser reports that the preference distribution has defied expectations by giving victory to Geoff Brock – according to Brenton in comments by 9987 votes (51.7 per cent) to Terry Boylan’s 9322 (48.3 per cent). Evidently those Nationals preferences were kinder to Brock than scrutineers believed.
Wednesday, January 21
11pm. Antony Green in comments: “The Labor scrutineers have been watching National preferences all week to work out where they are going. They’re flowing to the Liberals, which is why everyone’s given up on Brock closing the gap. Once the Liberals get half of the National preferences, there aren’t enough votes left to get Brock ahead of Labor.”
4pm. Based on Antony’s feedback, I have changed the minor party preference estimates as follows. Nats: Brock 45, Liberal 40, Labor 15. Greens: Labor 50, Brock 30, Liberal 20. One Nation: Liberal 50, Brock 30, Labor 20. That leaves Brock in third place, 1.2 per cent behind Labor.
3pm. With the addition of 3288 pre-poll votes, only a handful of postal votes remain to complete the primary vote count. These have made things interesting: coming mostly from Port Pirie, where the main pre-poll booth was located, they have split 1094 (33.9 per cent) to Brock, 1033 (32.0 per cent) to Labor, 868 (26.9 per cent) to Liberal, 179 (5.3 per cent) to the Nationals), 50 (1.5 per cent) to the Greens and 14 (0.4 per cent) to One Nation. Brock’s primary vote deficit against Labor has narrowed from 3.3 per cent to 2.5 per cent and, if my preference estimate is correct, he will just barely edge ahead of Labor on preferences and ultimately win the seat. BUT – please read this before commenting – these estimates are completely unscientific (see my 8.16pm entry from Saturday) and are evidently different from the calculations of Antony Green, who has spoken to scrutineers. He says: “Brock could yet pull ahead narrowly and win on Labor preferences, but it would require stronger flows of preferences to him from the National and Greens candidates than I think can be delivered. Not impossible but I would say it is unlikely.”
Tuesday, January 20
12.30pm. Antony Green has added 1795 postal votes which aren’t yet appearing on the SEO site, and they are very encouraging for the Liberals. Only 189 (10.5%) are for Brock, whose total vote has fallen from 23.1 per cent to 21.7 per cent, increasing his deficit against Labor from 2.0 per cent to 3.3 per cent. However, Antony notes that the 3000 pre-poll votes remaining to be counted mostly come from Port Pirie, which might at least staunch the flow. Terry Boylan has received 925 votes (51.5 per cent), increasing his vote from 40.2 per cent to 41.5 per cent and perhaps increasing his slim hope of winning even if Brock overtakes Labor. My table now includes a section for provisional votes, with a “votes counted” figure based on an educated guess that the final total will be 4500. Note that the preference projection now has Brock finishing in third place.
Monday, January 19
My general overview of the situation can be read at Crikey. Dovif in comments: “As for the scrutineers, the ALP will be trying to kick out as many ALP 1s as possible, while the Libs will be trying to increase the ALP vote. That would be fun to watch.”
Sunday, January 18
The Advertiser reports the Liberals are “confident” of retaining the seat, while conceding a “slight possibility” of defeat. The report says “almost 5000” postal and early votes were cast by Friday.
Saturday, January 17
9.00pm. I have evidently not been giving enough weight to the possibility that Brock will fail to get ahead of Labor. He trails by 2 per cent on the primary vote, which he would be able to close on preferences – but as Antony Green points out, independents traditionally do poorly on pre-poll and postal votes and the primary vote gap can be expected to widen. Antony deems it unlikely that the Liberals can win if Brock stays ahead.
8.16pm. That’s us done for the evening, with the result still up in the air. My preference estimate has Brock leading 7208 to 6837. I have distributed the minor players as follows: Nats: Brock 60, Liberal 30, Labor 10. Greens: Labor 55, Brock 35, Liberal 10. One Nation: Liberal 55, Brock 35, Labor 10. I have then taken the Labor vote, including those votes Labor received as preferences from the aforementioned, and given 80 per cent to Brock and 20 per cent to the Liberals. It was reported on Wednesday there had been 1700 early votes and 2200 postal applications, which can be expected to favour the Liberals quite solidly. Stay tuned over the next week or two.
8.11pm. Clare has indeed given Liberal candidate Terry Boylan the result he needed – 59.0 per cent (though down 7.9 per cent from 2006) against only 6.2 per cent for Brock.
7.49pm. Port Broughton and Tarlee now added – relatively good results for the Liberals, bringing my margin estimate below 5 per cent. If Clare can cut that further, the result will be truly up in the air.
7.47pm. Port Broughton has kind of reported, but the SEO is having more of those data entry issues (Brock on zero).
7.44pm. Just taking my first look at Antony Green’s site – his assessment is about the same as mine.
7.42pm. Still to come: Clare (2432 votes in 2006), Port Broughton (good Liberal booth, 849 votes in 2006) and Tarlee (259 votes). The Liberals will need very good results here, a good show on the many outstanding declaration votes and better preferences than I’m crediting them with.
7.40pm. Port Pirie booth of Solomontown gives Brock a slightly below par 35.4 per cent. The Liberals will be hoping for a big result in the very large country booth of Clare.
7.35pm. Three rural booths plus Port Pirie West now in – another plus 40 per cent result for Brock in the latter. My preference calculation now has him opening up his lead, so my summation from three entries ago may have been askew.
7.33pm. These are my preference estimates – would be interested if anyone disagrees. Nats: Brock 55, Liberal 35, Labor 10. Greens: Labor 55, Brock 35, Liberal 10. One Nation: Liberal 55, Brock 35, Labor 10. Labor: Brock 80, Liberal 20.
7.31pm. Unfortunately, the SEO is doing an irrelevant Liberal-versus-Labor preference count. Brock will clearly finish ahead of Labor.
7.30pm. Here’s roughly how I see it. Frome is evenly divided between Port Pirie and the rural remainder – the former is breaking 66-34 to Brock over the Liberals, and the latter’s doing the opposite. That suggests it should be very close, but this is based on my very rough preference guesses which if anything probably flatter for the Liberals. The locally knowledgeable Michael Gorey is calling it for Brock in comments.
7.28pm. Crystal Brook (rural) and Port Pirie South both in, another 40 per cent result for Brock in the latter.
7.21pm. Risdon Park South replicates Risdon Park East, with Brock’s primary vote around 40 per cent – my slapdash preference calculation now has him in front.
7.19pm. Three more booths in including a very exciting result for Brock in the Port Pirie booth of Risdon Park East – assuming it’s not a glitch, because the SEO has no percentage figures next to the raw results.
7.12pm. 2CP error corrected.
7.10pm. Five more booths in, including the first from Port Pirie – which Geoff Brock narrowly won ahead of Labor. That shuts out any notion of Brock failing to pass the Nationals, and could yet make things very interesting as more Port Pirie booths come in. Apologies for the 2CP error in the table – will get to work on it.
6.55pm. I’ve now removed Brinkworth’s alleged 14 Labor votes from my count.
6.53pm. Some explanations about the table. The “3CP” result assumes the last three standing will be Labor, Liberal and Brock, although Brock is well behind the Nationals on the basis of small rural booths. The “count” figure has been devised so it will add up to 100 per cent when all votes are in, whereas other media normally just show you the number of votes counted divided by number of enrolled voters.
6.50pm. Two more small rural booths, Brinkworth and Manoora, now in – although something’s obviously gone awry with Brinkworth, which has 14 votes for Labor and nothing in any other column, including the total.
6.39pm. As Judith Barnes notes in comments, the absentee vote could be over 20 per cent.
6.37pm. Two country booths reporting, Georgetown and Lochiel – excuse the mess in the Port Pirie entries in the table, it will correct when I have figures in. Only a small amount counted, but Geoff Brock might have hoped for more, remembering of course that Port Pirie is his stronghold. In noting the drop in the Liberal vote, it needs to be remembered there was no Nationals candidate last time.
6.15pm. Please excuse the messiness in the table above – I’m still sorting it out. The numbers there are test results rather than real figures.
6.00pm. Polls close. Official results here. First figures should start to come in around 6.30pm, by which time I should have my act together with my results table.
dovif, 2pp preference flows from last election are used for polling. Actual flows are used for predicting the result of an actual election, per Antony’s post above.
i hope they work it out as soon as possible, it’s getting bluddy uncomfortable here sitting with everything crossed for Brock.
http://www.crikey.com.au/Politics/20090119-Frome-by-election-Can-Brock-close-the-gap.html
Thanks for the article William.
Actually, its just struck me that this by-election actually has a lot in common with the 3-way contest in Mitchell at the last state election, where Liberal preferences helped get the independent candidate Kris Hanna over the line.
Primaries (%):
Mitchell: Clancy (ALP) v Hanna (Ind) v Gaffey (Lib): 41.0 v 24.6 v 20.7
Frome: Boylan (Lib) v Rohde (ALP) v Brock (Ind) 40.2 v 25.1 v 23.1
Strikingly similar …
The difference being that Mitchell was ALP held, ind assisted by Lib prefs, while Frome is/was Lib held, ind assisted by ALP prefs, but both inds are ALP-leaning.
The parties will always prefer the seat to fall to an ind rather than the opposing party.
In rural areas the Green vote is much less of an ideologically left vote than it is in the cities. Labor will make up little ground on Brock out of Green preferences if any.
IMHO the Greens chose not to direct any prefs because they wanted Brock to get up like Labor does, and knew Brock was the only realistic chance of taking the seat from the Liberals, and preferencing Labor as their base would most likely have preferred would only have hampered Brocks chances.
I still maintain Family First kept out of the contest by choice or by Liberal request to prevent any leakage of right-wing votes to Brock. MHS likes to claim most of the swing the Libs suffered was due to the Nats running, but then a) he slams them for being an offshoot of the Labor Party, and b) most of the 5% Family First vote would have gone back in to the Liberal fold, reducing the drop in their primary vote.
Bob, I don’t think you’re right. If the Greens wanted to elect Brock, they should have lodged a how-to-vote to go on the voting screens that would have said that. The Greens lodged a voting ticket with second preference to Brock, then National 3, and split 50:50 between Labor and Liberal. But that voting ticket will in the end only have mattered to save a dozen or so votes with incomplete preferences. By having no how-to-vote on the voting screens, they have probably reduced the number of preferences flowing to Brock, not made it easier for him to win.
My point is that the Green base would have been split over whether to pref Labor or Brock second, so to avoid stepping on any toes they did not release a how to vote card, and let people make their own decision.
Unless the South Australian Greens constitution is different from other states, it is local members who make the decision on preferences, so its a bit hard to know whose toes the members were treading on if they’d made a decision to direct preferences. By not directing preferences they have ensured their preferences were split, whereas if they had directed preferences they may have split their base but would have delivered a stronger flow of preferences to Brock. I presume by saying ‘IMHO’ you are trying to work out the Green strategy, whereas I’m pointing out that if the Greens were trying to do what you suggest, they’ve gone the wrong way about it. I think you’re making the mistake of trying to think strategy. The Greens interest in the campiagn was about carbon emissions from the Port Pirie smelter, and if you look at the policies of the candidates, I think you can see why the Greens didn’t direct preferences.
Good point.
Antony, I agree that the Greens chose to abstain from directing preferences to Brock and the ALP because of their policies on smelting. I understand completely what you wrote (post 96) about the possibility of ALP preferences pushing Brock over the line past the Liberals, but the ‘other’ ABC post is still showing a progressive count after preferences (which includes, presumably, a counting of Brock preferences) with the Libs retaining the seat.
The overall effect of Brock’s standing seems to have been to split the ALP vote, dramatically. If the ALP preferenced Brock over the Greens because of the smelter, then they should consider that the Greens still managed to pick up a +0.6% swing in an electorate that has relied strongly on smelting. This in itself should suggest a shift in public mood.
Incidentally bob1234 mentioned a swing of +0.8% to the ALP, which the ABC is now showing on its website.
+0.6% is a shift in public mood?
Are you forgetting the Democrats didn’t stand a candidate, after receiving 2% at the 2006 election?
0.6% is white noise.
freihans@112 – the ‘other’ ABC post is showing the indicative preference count between Labor and Liberal candidates conducted by the South Australian Electoral Commission. If Brock finishes third during the distribution of preferences, this indicative preference count will be the best guide to the outcome of the election. If Brock finishes in second place, then the indicative preference count will be of no use. At this stage, it is not clear whether Brock will finish second or third.
Indicative preference counts are undertaken by Electoral Commissions so that observers will have a better indication of who will win a seat after the distribution of preferences. Indicative counts are done by the Electoral Commissioner nominating beforehand who will be the final two candidate for the indicative preference count. In Frome on Saturday niight, the Deputy Returning Officer in each polling place would have opened a letter after the close of polls telling them who were the two candidates for the indicative count. In Frome the candidate were Labor and Liberal. After the tally of primary votes for all six candidates, the ballot papers of the other four candidates were re-examined. For each ballot paper of these four candidates, a determination was made as to whether the Liberal or Labor candidates received the ballot paper’s preferences. These prefences were tallied along with the primary votes for the Labor and Liberal candidates to produce the indicative preference count.
In 95% of cases, the Electoral Commissioner picks the correct two candidates. In some cases, the wrong candidates are chosen. Frome may be one of those cases. Two days after the election, we are still not sure who will finish second, so it is not surprising that the Commissioner was unable to make this determination before election day.
In the official distribution of preferences, conducted at the end of this week after all votes have been counted, a very different procedure is undertaken. From the primary vote tallies, the candidate with the lowest total is excluded, and the candidate with the next available preference for each ballot paper is determined. In the case of Frome, the first candidate to be excluded will be the One Nation candidate, followed by the Green candidate, then the National candidate. At this point only three candidates will remain, Liberal, Labor and Brock. If at this point Brock is in third place, he will be excluded and the Liberal candidate will be elected, as was indicated by the indicative preference count. If the Labor candidate is in third place, he will be excluded and it is most likely that Brock would then defeat the Liberal candidate.
We won’t have a better handle on whether Brock can reach second place until the pre-poll and postal votes have been counted. On the indicative count, the Liberal candidate will win if Brock finishes third. But until the rest of the votes are counted, and the official preference count undertakne, we won’t know if Brock finishes second or third.
freihans@112 – I told Bob that the swing would go to +0.8% this afternoon because I made that calculation and typeset the page. The swing on election night compared the totals we received with the election night percentages from 2006 and this showed no swing. Yesterday an extra 200 or so election night votes were included, and these new election night tallies compared to 2006 showed a swing of 0.3% to Labor.
I have increased the swing to 0.8% because the current totals are now being compared to the overall results from 2006. So where on election night the ABC site was publishing a prediction of the final outcome by comparing election night 2009 with election night 2006, the site is now comparing election nigght 2009 with overall figures from 2006. So on election night the swing took account of the 2006 trend in declaration votes, while the new swing of 0.8% takes no account of this trend. However, if the declaration votes fall in the same ratios as in 2006, then this 0.8% swing will now begin to fall toward the 0.3% swing comparing election nights.
I know this sounds complex, but it is the only way to compare election results.
I’ve just discovered if I mention someone elses name, my comment goes into moderation. Comment @112 has response currently in moderation with a complete explanation of how preference counts are conducted on election night.
Hey, its a smelter town. The Greens should have barely registered at all – but they actually picked up an increase. However small that increase is, it matters because Green parties all over the world seem to show a trend for gradually gaining in popularity. It is also relevant because smelting could move from Frome to Hobart, meaning a growing environmental consciousness would be able to flourish without competing with local industry.
The creation of the Climate Change Coalition party in 2006 reflects this too. The CCC ran in the Hunter Valley at the last Federal election, and did surprisingly well considering over half the people of the Hunter Valley live in centres that focus on minerals and heavy industry for income. This could be because reliance on minerals, except as a transport hub, is being phased out in Newcastle and public sentiment is changing accordingly.
Could be interesting, in the long term.
Thanks for the explanation Antony. It’s a wonder isn’t it, a party can register a sizable decrease in its primary vote but still be shown to have a swing in its favour while another party can register an eensy little swing in its primary vote that becomes significant when industrial dynamics are considered.
Considering a government has never won a seat in a by-election from an opposition in the history of the South Australian parliament, I wonder what the closest margin has been, and if a seat held by an opposition on a margin as low as 3.4% swung toward the government?
Easy answer to that. Frome by-election 1960. Swing to government was about 5% and the final margin for the opposition just 17 votes. You can read about it at http://www.abc.net.au/elections/sa/2009/frome/
Well there you go!
Bloody Frome.
I wouldn’t think Greens issuing a How to Vote card would have made much difference. Haven’t checked but doubt if many booths had Greens handing out. The HTV cards put up in booths have virtually no impact because very few people read them. Though a few more read them than the Senate HTV tickets which are pinned up on a wall somewhere and I’ve never seen anyone read them before voting. And Green voters are some of the least likely to strictly follow a HTV card. Same comments apply for the few One Nation votes – these will be really random votes and I would expect them to go just slightly to Brock above Libs and Labor because dislike of major parties is about all that some ON voters will think about. Virtually none of them will have had a clue about what the ON HTV card said. Also doubt that FF not having a candidate makes any difference – voters inclined to give someone eg Brock their second pref would mostly give him a first if no FF candidate. To say otherwise assumes thought processes which are odd. People who “protest vote” for minor parties generally do so in a conscious way.
the lib candidate isnt crowing so loudly right now.
http://www.news.com.au/adelaidenow/story/0,22606,24935050-5006301,00.html
[ SUPPORT for the Liberals in the Frome by-election has eroded in a recount – lifting the chances the party may lose a seat held since 1970. ]
Held since 1970? Not bad, for a seat only created in 1993. It may be referring to Stuart (previously contained Port Pirie, won by Graham Gunn in 1970), but it’s still a weird way to word it.
In South-West Coast at the 2006 Victorian poll the Greens had an open ticket and only staffed some big urban booths. Green preferences did not go very strongly to Labor. The Nats had a ticket that gave preferences to the Libs but failed to staff many booths and their preferences drifted notably. In a rural seat you would expect a strong independent to do well out of minor party preferences. I would favour Brock at this stage. Monday’s Advertiser story was lazy journalism 101 couldn’t the reporter have asked the Libs to explain their prediction?
I still don’t get why there was a 2PP swing to Labor considering the slashing of rural health services in the electorally meaningless rural areas.
So how’s this for a theory?
The voters of Frome, where Kerin had a personal vote, rebelled against the Liberal leader Hamilton-Smith for destabilising his leadership prior to the 2006 election, where he half-challenged Kerin for the Liberal leadership just a few months prior to the 2006 election?
Updated figures following counting of postal votes at http://www.abc.net.au/elections/sa/2009/frome/
Bugger Antony, bugger.
What votes are left from here? What margin do you predict now, and between who?
Antony, am I missing something, or are these votes not on the SEO site? Subtracting what I have already from your figures, the postal votes are as follows with the polling booth results in brackets:
LIB 925 51.5% (40.2%)
ALP 436 24.3% (25.1%)
IND 189 10.5% (23.1%)
NAT 158 8.8% (6.5%)
GRN 67 3.7% (4.3%)
ONP 20 1.1% (0.7%)
As you can see, a dismal set of figures for Brock.
Suprising, considering the Telegraph said a significant number of the postal votes were from Port Pirie.
I’ve written it all up here http://blogs.abc.net.au/antonygreen/
Looks like Labor are polling a little too well for Brock to have any hope of getting second place – hard to see anything other than a Liberal win now. I’m waiting for the calculations of how the votes need to break now for Brock – but its looking of theoretical interest only at this stage.
In addition, I’m hardly surprised by the postal vote scenario – loads of farmers go on holidays in SA during January – and they sure aren’t going to be voting Labor in any great number.
I know Antony predicted that Brock would do poorly on postal votes but a drop from 23% booth count compared with 10.5% postal is enormous. It’s a great lesson to anyone thinking of running as an Independent. A lack of organisation and strategy by Brock looks like losing him a seat he could have won.
I agree to some extent Diogenes, but I guess it depends where the postals have come from – if there are a significant number away from the Port Pirie area, the drop probably isn’t that huge. I know there has been talk of large numbers of postals coming from Port Pirie, but the reports may well be confusing these with pre-polls. In any case, the numbers aren’t flash for Brock whichever way it is in reality.
Judith might get a kick out of the exchange between myself (fascinated of adelaide), and dj of adelaide @ http://www.news.com.au/adelaidenow/comments/0,22638,24935050-2682,00.html
I’m waiting for my post to be approved. Honestly, dj really does not understand primary and 2pp.
Approved!
Well the ABC progressive count after preferences swings seem to reflect the truth a bit better now.
Antony, do the total votes (16,321) include the 3,000 pre-poll votes. Have 73.7% of enrolled voters voted, or 87.2%?
f@139 – you’re still getting the numbers round the wrong way. What has happened now is that the primary votes are now lining up in such a way that it has become clear the result will be the progressive two-party preferred count. The ‘ABC progressive count’ is the Electoral Commission’s preference count. The total votes does not include the 3,000 or so pre-poll votes because the pre-poll votes haven’t been counted. The current count of votes represents is 73.7% of the enrolled voters. After the pre-poll are counted it will be about an 87% turnout, which is not bad for a January by-election.
http://www.news.com.au/adelaidenow/comments/0,22638,24935050-2682,00.html
http://www.news.com.au/adelaidenow/comments/0,22638,24937613-2682,00.html
It is hilarious to see the desperate Liberal supporters claiming Labor has been slammed in this by-election. There was barely any swing!
I wouldn’t be too worried Bob – its just another day in the office for the Liberal party intelligentia.
Bob i dunno who was the funniest, dj or Andrew, i couldnt resist putting in my two bits worth though i dont think it’ll be published, mine never usually are,i’ve copied it below, your a norty boy Bob teasing the minnows lol.
Andrew of Unley, i’ve been amused by your stoushing with Fasinated but i think maybe, just maybe you’d better stop writing about things you dont understand, egg on ones face doesnt help with the complexion, i just happen to know that Fascinated not only knows full well what he’s writing about but also has access to accredited professionals in this field.
tutt, tutt, Fascinated, you cant force teach anyone lol, sorry to spoil your fun. J.B.
bob
I couldn’t describe myself as Fascinated so I went with Interested. There’s an awful lot of desperate spin from the Libs isn’t there.
Antony, the ABC is showing two party swings that are less incongruous with the primary vote swings shown yesterday. I think this is what we both expected, deep down, although the Brock ‘suspense’ story made statistics interesting for this time of year. Good work.
I thought 87.2% was a fairly reasonable turnout too, which is why I asked, just to be certain. I suppose we will never know who the other 12.8% would have voted for, or what would have happened if the Climate Change Coalition and a few other preference directing candidates had stood.
And it continues at AdelaideNow. At least we have conclusive proof that Liberals simply lack any understanding of the electoral system.
whoops, back to see what interested had to say, i just posred under the nick Judy–how boring is that!
Me: David, if next year there’s a 16% primary and 1% two-party swing away from Labor, they’d still get back in with a thumping majority, losing only one seat! Sorry to rain on your parade.
And the charades continue. What fools.
That is all we need in South Australia, a former copper in State Parliament! The next series of Law and Order may as well be filmed in the wowser state! The Parliament is chockers with religious, law and order and anti-gambling fanatics! I cannot believe how unrepresentatative our parliament has become! The conservatives elites in the Labor and Liberal Parties are ruthless of allowing any form of diversity in their respective parties. Add Family First and the nutters plus in overflow to Xenophon and the parliament is fast becoming a huge farce, if it wasnt so serious!