The latest semi-monthly Reuters Poll Trend figure, a weighted composite of results from Morgan, Newspoll and ACNielsen, continues the gentle trend back to the Coalition that has been evident since May. On the primary vote, Labor is down from 47.7 per cent to 46.9 per cent and the Coalition up from 39.5 per cent to 40.4 per cent. Kevin Rudd’s preferred prime minister rating is steady on 46.4 per cent, while John Howard’s is down from 40.5 per cent to 40.2 per cent.
[This shows that there are some issues where he IS prepared to differ. It’s a smart issue to differ on, because it will help get Green preferences.]
And it is an easy argument to make. The fact we have so much uranium is what gives us influence in the aim of stopping the production of nuclear weapons, by selling uranium to countries producing and testing weapons, then we have just given up our best bargaining asset.
Tony at 246
I think Rudd is being quite clever in the way he is framing the narrative of both his agreements and disagreements with Howard.Take the India issue for instance, the ALP has a big history of nuclear non-proliferation…. the Canberra Commission was their baby.So by disagreeing with Howard over the Indian deal, not only does it play well to the Greens, but it also has an historical ALP narrative to claim consistency with which helps with the quality end of the beltway commentary.
Rudd has done many things – but complete backflips with a double twist and a pike isn’t one of them.He is, IMHO, quite intelligently crafting out differences and agreements with a pretty well constructed historical narrative and consistent chronology.
{Well, I reckon First week of December, there is CHOGM in Uganda and Howard will want to be there for one last photo-op with his Commonwealth Buddies.}
Frank, Howard will have to get the Photographer to do a cut and paste job with Bush though.
And check where the swings to the Coalition were in Victoria in 2004. A range of outer suburban seats, where interest rates were big. And above all, a band of eastern suburban electorates where the government ran a very strong campaign against the Scoresby Tollway. Coming from Sydney, where we have nine tollways, and where the Federal government thought the Western Orbital should be built as a tollway, I always thought it was generous and indeed magnanimous that the Federal government was prepared to offer so much money raised from taxes across Australia, just so that people using the Scoresby tollway wouldn’t have their pleasurable ride interrupted by the ker-ching of their e-tags. But will this issue have the same salience in 2007?
That’s right Possum. Living in Victoria, I see no reason why this state should deliver less of a swing to Labor than anywhere else. The State Government is popular, and has managed a leadership change extremely well. The State Opposition is making little impact on anything. The state economy is very sound, house prices are rising steadily, unemployment is low. There are no particular issues like council amalgamations that would sway Fderal voters. I would expect Victoria to match any national swing to Rudd, or even exceed it a bit.
Victoria has been more friendly to Labor than most states in recent federal elections (I don’t regard 1990 as recent!). So I see a 6% swing as very achievable. Labor has just changed its candidate in Latrobe, aand is putting extra resources into a couple of other marginal Liberal seats.
I’ve maintained for a while that Corangamite is a seat ripe for a Labor win, and the sitting Liberal member is getting beyond his use-by date. Fran Bailey’s seat of McEwen is creeping down more and more into Melbourne’s mortgage belt, where any personal vote she might have in central Vic will mean buggerall. Laatrobe is very winnable for Labor. I have some doubts about McMillan, though, as the area swung against Labor in the last state election, over water issues, I suppose.
Antony:
How will the 2005 WA State Election results translate to votes in Stirling & Hasluck. ?
And do you think the Brickworks debacle in Hasluck will lose votes for Stuart Henry ?
(BTW, I’m in the Neighbouring electorate of Pearce)
{Rudd has done many things – but complete backflips with a double twist and a pike isn’t one of them. He is, IMHO, quite intelligently crafting out differences and agreements with a pretty well constructed historical narrative and consistent chronology.}
Possum, the desperation being displayed by Howard at present, demonstrates clearly that he doesn’t underestimate Rudd but his main problem is he has no effective counter against him.
When Rudd said that he was going to “mess with Howard’s mind”, he really meant it and he is setting Howard up consistently using the “master tacticians” own tactics against him. More to come yet.
Nostradamus is looking at uniform swings.
Don’t you understand that the swing against the government is in a lot of different areas.
(1) Those pissed off with the effects of workchoices.
(2) Those with a mortgage
(3) Those fed up with the lies.
I have no doubt that seats with current double digit majorities are going to fall simply because of the above.
Sir Eggo,
If the Newspoll quarterly breakdowns are to be believed (and I see no reason why they would be grossly inaccurate) – there’s strange things happening in NSW (actually there’s strange things happening all over the bloody place). There’s a 12.2% swing against the coalition in NSW which seems to be intimately intertwined with the 14.6% swing against the Coalition safe seats. Even if we assume that level of swing halves by the election but the pattern of the swing holds, that delivers more seats to the ALP than the pendulums suggest because the swing would be seriously non-uniform in NSW that advantage the ALP (on average happening more in marginal and semi-marginal Coalition seats than the ALP safe seats which by the same Newspoll data are only registering a 4.1% swing towards them). There hasn’t been a quarterly Newspoll in in the Howard government period which has suggested such dramatic non-uniform swings occuring.This will be an very very very interesting election if the weird patterns in the Newspoll quarterly breakdowns hold, even if they dont hold their curent level.
Antony Green, I’m curious as to whether you have a prediction on this election? I’d like to know how an electoral-guru thinks things are going…
[Based on the 2004 results, this will mean Rudd needs 53-54% TPP.]
Mumble posted about this:
http://www.mumble.com.au/federal/2ppmargins.htm
It seems that around 53 2pp is what results in a change of government; even if the party that forms government needed less on uniform swings.
1972, ALP needed 51.1, but got 52.7
1984, ALP needed 51.0, but got 53.2
1996, Coalition needed 49.7, but got 53.6
In theory rudd could win by a seat or two with 51 or 52 2pp, but in practice he needs somewhere between 52 and 53. He would win in a landslide if he got 54, the election calculator says he would get 87 seats:
http://www.ozpolitics.info/guide/elections/fed2007/election-calculator/
If the swing is on, it is on, and he will get 53.
Antony, the Scoresby freeway/tollway issue is not an issue any more in Victoria. I think the vast eastern suburbs of Melbourne will vote on the standard issues – state of the economy, interest rates/housing and rents, the “it’s time” factor, and WorkChoices.
Nostradamus does make a fair point, that Victoria is already Labor’s strongest state, nd a further big swing might be hard to achieve. But the state Labor government hasn’t had any troouble getting massive swings.
BTW Nostradamus, Ballarat needs a bigger swing than you suggest to be won by the Libs, and they won’t achieve it. Catherine King is a strong Labor member, the area now votes solidly Labor in state elections, and it’s the home town of Steve Bracks. With a strong candidate and some good local policies, the Libs would have a show of winning Bendigo, but their heart doesn’t appear to be in it. Recently, a French-owned factory in Bendigo failed to win a contract to build Busmaster army vehicles for the US. Labor is claiming it’s because of the US not liking the French opposition to the Iraq war (!), and that the contract went to US companies, thus costing many potential jobs in Bendigo.
I reckon it’s drawing a pretty long bow myself, but it certainly won’t help the coalition in Bendigo.
There is a potential barrel of pork though. Both Bendigo and Ballarat are short of water, and the Vic Government has pledged to build a pipeline from north-east Victoria to supply them, and that’s unpopular in some areas. A major promise about water from John Howard might help get votes in Bendigo, in particular, and also provoke a fight with the Brumby Government. But I suspect Lib internal polling is showing both Ballarat and Bendigo as pretty safe for Labor, and not worth much effort.
Comparing state and federal results is rarely worth the effort. And i’m in no position to have an opinion on the importance of the Brickworks issue.
I’ll also say this about Victoria. Remember that Victoria has historically been Labor’s weakest state. That position has reversed since 1980, with the notable exception of 1990. Labor got only 50.3% in 1996, then the next worse since 1980 was 51% in 2004. The main reason why Labor has a better record at Federal elections since 1980 than at any time before 1980 has been its vastly improved performance in Victoria. People forget just how pathetically left-wing the party was until Federal intervention in 1970. I’d also say South Australia has been out of line with its traditional Labor vote since 1993, and polls seem to indicate that Mike Rann has renovated the Labor brand in South Australia. Polls would indicate South Australia will deliver a minimum 3 seats.
I am strongly of the view that the current margins in many seats is inflated. All the 2001 marginals swung heavily against Labor in 2004. Was the 2004 swing on those seats the sign of a re-alignment, or just a reaction to the 2004 campaign? If it was just a reaction based on Latham and interest rates, then there are a lot more seats in play then just the current 16 most marginal everyone keeps looking at.
[Possum, the desperation being displayed by Howard at present, demonstrates clearly that he doesn’t underestimate Rudd but his main problem is he has no effective counter against him.]
Exactly, which is why Howard is trying to pretend that Peter Beattie is Leader of the Opposition.
I agree with Antony re the South Australian analysis. When I looked at state v federal alp primary voting based on the closest state/federal election recently, there is a reasonable correlation apart from the 1993 state election, which was an abberation due to the state bank disaster. If the ALP can achieve anything even in the same ball-park as the 2006 state election, Boothby and Sturt will be in real strife.
Antony at 254
Interesting what you say about the outer suburban interest rate swings to the Coalition in 2004.I’ve been pouring over the George Megalogenis (sorry for the spelling George) “Meganomics” data over the last few days and nearly all of the seats with the heavy mortgage burdens seem to be swinging, but in really non-uniform ways.The data is really clouded because of the lack of homogeneity in the economics of the seats and the way the mortgage burden gets polluted in determining the swing data by Non English Speaking Background demographics to name but one issue (which is a bit strange, I know… there’s been much frustration over that 😉 ) – but I can see that swing there, I just can’t quite pull it out of the available data properly because of the inherent aggregation bias contained within the data.
If the ALP swings in the heavy mortgage burden areas are happening the way some really experimental (and let’s be honest – quite econometrically dubious 😉 ) measures I’ve constructed as a means to clear up the data are suggesting – this election will, if the Newspoll data holds (or even If the pattern of the Newspoll data holds) have quite a few unexpected ‘bolter’ seats.
[Well I say bring on fixed terms so we political junkies can organise our diaries and schedules well in advance!]
Yes please! I wish we could vote for local, state, and federal on the same day, every four years.
248 Simon Howson
Agree with both those comments. The ‘right’ of the executive to unilaterally call an early election is one of the great weaknesses in our political system, that is scammed at every opportunity by both sides of politics. The only person who should be able to bring forward an election date is the Governor General.
[I am strongly of the view that the current margins in many seats is inflated.]
If you go by the 2004 pendulum, ALP would get 16 seats on a 3.7% swing. 17 on 4.3, and 18 on 4.7.
This takes into account that they won Richmond, Hindmarsh, Adelaide and Parramatta at the last election.
Adam,
A lot of the time they a selling short
Possum, care to predict any bolters? I remember the NSW seat of Hughes was an absolute bolter for Howard in 1996, dumping Robert Tickner into oblivion. Jackie Kelly’s win was a similar result.
It tends to be mortgage belt seats that deliver shock results, I think.
[The ‘right’ of the executive to unilaterally call an early election is one of the great weaknesses in our political system,]
Along with party discipline, the executive being part the the legislature, compulsory preferential voting, and the head of state being born into the job? 😛
I wouldn’t mind fixed 3 year terms, 8 years seems like a really long time for a senator, 6 is more reasonable. Having all the elections for all levels on the same day is I think, more important. It could be on a public holiday Monday to create another long weekend. 😛
More importantly, I’d do away with compulsory preferential and compulsory voting (well, it is actually compulsory attendance, because they can’t actually tell if you just left the ballot blank, or otherwise made it informal). I think that is an afront to a basic right in any democracy; the right to not care.
Tony – I’m not brave enough to really predict any bolters yet according to the data because I’m not truly satisfied that my data analysis is rock solid at the moment.But I’ll certainly have them up before the election, even if I cant get them any better and the election is called.
But it must be… HAS to be a collection of seats that the Coalition currently holds between margins of 6 and about 9%, and a fair whack of them will be in non-capital city areas with median incomes around the 4th income decile and which are currently held by Liberal rather than National members, and probably in NSW or Qld.
From on the ground experience, I think Petrie is a bolter.If Rudd announced funding for building the Redcliffe rail extension – he would even further make this seat into safe ALP.
Possum, I think George is over-analysing the data a bit, especially when he starts to look at the age break-up. Compared to other countries, Australian electorates are relatively homogenious. There is a clear city/country divide. In the city, there are clearly affluent Liberal voting electorates, and equally clearly less affluent and heavily migrant populated Labor electorates. But nothing like the startling class and regional difference you see in British electorates. I’ve seen the vast models they use in the UK to call elections based on census variables. They have to do that, because in the UK you do not get progressive results in a seat, merely the final read out of the result in each district. In the UK, election night is a process of projecting the results you have on to electorates where you have no result, and combining them with any exit poll data. (Think the famous monty python sketch for those old enough. http://www.youtube.com/watch?v=rooOWZnFYkI)
Here the swing from seat to seat is relatively uniform, especially if you take account of the state, and factor in retiring Mp factors. The 1998 and 2004 elections were interesting because there was another factor at play wokring on the swing, and that was clearly mortgage belt seats. At both elections, the swing was quite different in the mortgage belt, not swinging back to Labor in 1998, and clearly swinging away from Labor in 2004. That’s why Macarthur and Aston are now safer seats than North Sydney and Kooyong. But are those safe mortgage belt seats realy as safe, or can they swing more having swung so far to become safe? Remember that in 1996, seats like Lindsay, Dobell and Hughes were way bigger than the National swing.
The quality of the data in Australia on election night is so good, that media organisations don’t bother with exit polls. Except for Sky News, who have to have something to talk about given they start their election coverage an hour before the polls close.
A bit OT I know, but this sort of desperation can’t help the Coalition pull back the gap in the polls, surely.
{FOREIGN backpackers are being hired by the Federal Government to check the legality of the controversial Australian Workplace Agreements (AWAs), Labor says.
Opposition industrial relations spokeswoman Julia Gillard said in order to apply the “fairness test” on a backlog of 100,000 AWAs, the Government has employed about 200 temporary workers.
The workers are given just six days training before starting the checking process, she said.}
http://www.news.com.au/story/0,23599,22256836-29277,00.html
Simon, no I mean on 2001 election results. The Coalition had 22 seats under 4% going into the 2004 election, and you’ll find a bunch of them now all sit up on 6-7%. LaTrobe, Longman, Deakin, Dobell. That’s what I mean by the 2004 election may have left seats like that with inflated margins. All the Coalition’s marginals going into the 2004 election got huge boosts from the 2004 result. Was that swing peculiar to the 2004 election, or did it indicate an underlying shift? Seats like Petrie were very marginal in both 1998 and even 2001, but the margin shot up in 2004. A sign of an emerging Liberal dominance or a campaign effect? I suspect the latter, unless every opinion poll this year has been wrong.
And this from John Symonds won’t help those “Howard Battlers” either.
[AUSTRALIANS should brace for some tough times ahead as problems continue to plague international money and mortgage markets, Aussie Home Loans chief John Symond says.
As the Australian sharemarket endured another tough day, with shares slumping 5.5 per cent at one point, Mr Symond said he expected current conditions to continue.
“It just reconfirmed in my own mind that these are volatile times,” he told the ABC’s The 7.30 Report.
“It’s not a rain shower. I think the clouds are forming and we’ve got to be in for a tough ride.”]
http://www.news.com.au/perthnow/story/0,21598,22257662-5005361,00.html
I know that the Liberal ostriches in here like to just ignore the evidence contrary to their desired worldview, but as has been pointed out on a number of occasions, 4 out of the last 5 Westpolls have shown a swing to the ALP in WA. Only 1 out of 5 has shown a swing to the government.
Again, things could change, but at the current moment the evidence is that the swing in WA is, like the rest of the country to the ALP, at a reduced strength, but still easily enough to at least threaten Stirling and Hasluck.
Possum, do you think Kalgoorlie could be a bolter? I notice Howard paid it a visit recently. Labor suffered last time because their candidate died I think just before the election.
While you might think the mining boom has brought great wealth to the seat, it’s also brought a lot of inconvenience, and housing is hugely expensive or unobtainable, particularly if you’re in the majority, and are not a miner. A lot of young people would have moved to the seat since the last election. And even a lot of rich miners still vote Labor. There’s also an aboriginal vote. Labor has a strong history in the seat.
Other random bolters, for which I only have gut feeling on which to base my prediction, could be Dickson (the backlash against Cheryl Kernot pushed Labor’s vote back beyond where it should be, and it’s a mortgage belt seat), and McEwen (a former rural seat now stretching into Melbourne’s suburbs). And what about Sturt, on the same rationale as I’ve used for Kalgoorlie (and because there’s clearly a strong swing on in SA)?
I may be completely wrong, but it’s fun to guess anyway.
I am strongly of the view that the current margins in many seats is inflated. All the 2001 marginals swung heavily against Labor in 2004. Was the 2004 swing on those seats the sign of a re-alignment, or just a reaction to the 2004 campaign? If it was just a reaction based on Latham and interest rates, then there are a lot more seats in play then just the current 16 most marginal everyone keeps looking at.
Thank you Antony. My homework for tomorrow will be looking for the 2001 marginals that swung heavily against Labor in 2004. My next question will how much influence did 9/11 have on the 2001 outcome cf to the Latham bullyboy ‘handshake’ at the radio station in 2004. There are some advantages in having the flu and being ordered to “stay away from other people” for a few days. That’s abit hard for a social worker to do so I stay home.
Then again I could just ask Possum or Aristotle to go to their data banks and punch out the name for me . hint hint. lol.
Antony – I havent been using the wisdom of George (and no offence to George), I’ve just been using his 2 Meganomics data sets that News Ltd paid to get pulled out of the census data.
http://www.theaustralian.news.com.au/files/meganomics1new.html
http://www.theaustralian.news.com.au/files/meganomics2new.html
And combined that with my own database of primarily Newspoll results, but also the other polling agencies results, individual seat polling that’s been reported and a few other bits and pieces.
The age cohorts have no statistical relationship to any of the available polling results when it comes to analysing the data by seat.In fact, most of the Meganomics data have no statistically significant relationships to what’s happening in the polling data.
I agree that we are generally socio-economically homogeneous (many Brisbane seats being the exception), but with issues like interest payments to disposable income (basic debt servicing obligations for mortgages for instance) that homogeneity breaks down in more seats than not. The distribution of mortgage holders, by area, by the ratio of their debt to real income levels is quite diverse.
And if we look at the better end of the polling data, there is likewise a big diversity in the swing to the ALP going on somewhere in at least three states (NSW, QLD and SA).The Meganomics data suggests that mortgage burdens are playing a serious part in that non-uniformity – I’m just not sure how because of the aggregation bias preventing further delving into the numbers.
Essentially I completely agree with you over “mortgage belt seats” being on an inflated margin for the Coalition – but some of those seats seem to be, nay, MUST be swinging far larger than other seats if the quarterly newspolls in conjunction with the latest census data are correct.
Interestingly, the percentage of homes with 4 bedrooms or more in a given seat seems to be seriously diminishing the swing that is happening to the ALP.The more of those McMansions in a seat, the smaller the swing to the ALP by any measure I look at it – be it experimental TPP estimates taken from my Pollycide series, through to more realistic numbers using linear programming to make those experimental TPP estimates (derived by using weighted newspoll breakdowns on State, marginal vs non marginal and capital vs non capital city seat breaddowns) behave to known swings… right through to using simple state based swings.
That’s why I think what is happening isnt as uniform as it usually is.The currently available data just gets indigestion with uniform swings.Maybe that will change by the election, but currently, non-uniformity is the new black in a lot of seats.
Has anyone seen the ninemsn poll its running 5 to 2 for pauline hanson beong right about about muslim immigrants
Tony, I wouldn’t be surprised if Petrie and Longman are in play, and possibly Dickson. Longman and Dickson are outer suburban seats although Dickson also has some conservative voting shires and Petrie has every chance of being won by the ALP. Petrie seems to be a seat usually won by the party that wins the election – the last time that didn’t happen was in 1984. Petrie is usually marginal.
Is that 5 votes to 2 votes?
[Has anyone seen the ninemsn poll its running 5 to 2 for pauline hanson beong right about about muslim immigrants]
Does this mean she’s figured out how to use a computer?
Adam
The trend is your friend. Sell today and buy tomorrow when it is cheaper.
Those who want a paper loss rather than a “real” loss can hold and the runners will end up in front (as long as the trend doesn’t turn)
SirEggo,
In terms of constructing a possible ALP win, there’s enough out there to show it is possible without heading into unprecedented territory.
I’ve done a few calculations based on the work of others (and apologies to them for the following hack summaries). You can take Simon Jackman’s pendulum and apply a proportionate state swing based on Newspoll. You can then discount for a more realistic TPP (in the following case 52.5% – which I also moderated so that bigger swings in particular states came down by more and smaller swings by less). Using Newspoll’s safe coalition/marginal/safe ALP swing data, you can then “force” coalition seats on more than 5% TPP by a further 2.5%. What that gives you is the following seats going to the ALP.
NSW – Parramata, Wentworth, Lindsay, Eden-Monaro, Bennelong, Dobell, Page, Paterson, Cowper, Robertson
NSW forced – Hughes, Gilmore
Vic forced – Deakin, McMillan, Corangamite, La Trobe
Qld – Bonner, Moreton, Blair
Qld forced – Herbert, Longman, Petrie, Flynn, Hinkler
WA – Hasluck, Stirling
WA forced – Kalgoorlie
SA – Kingston, Wakefiled, Makin, Boothby, Sturt
Tas – Braddon, Bass
Nt – Solomon
On this basis there are 35 seats within contention on a 52.5% TPP. On a smaller TPP there will still be plenty of seats in contention.
As I said at the start, this is a constructed scenario and will no doubt bear no resemblance to election night outcomes – changes to relative state swings, local factors and the all important TPP will be relvant to the final outcome. However, I think there are a couple of points which can still be taken from this analysis. First, in my view the ALP is alomst certain to win if it gets 52% TPP, and very likely to win if it gets 51.5%+ TPP. Below that is less certain but still very possible. Second, and leaving aside that I haven’t factored in any seat-specific matters, there is plenty of scope for the ALP to pick up a number of less obvious seats.
As a final point, I’m certain that both the ALP and Coalition are going to be focussing on 30-40 seats for this very reason.
Vote: Do you agree with Pauline Hanson on Muslim immigration?
Yes : 50128
No: 20689
I Voted No of course – but what’s the bet one of Pauline’s Computer literate friends organised a Bulk email garnering support 🙂 Or it could be people voting from “The Shire”
Tony – what I know about Kalgoorlie could fit in an eggcup 😉
From a mathematical perspective, Kalgoorlie could be in the mix as one of those seats in the cross hairs of a bigger swing than expected – it seems to fit the pre-requisites, but there’s lot of seats like that and not all of them will swing that way.
Although if Howard is paying it a visit, it might say something.I’ve been surprised at the seats Howard has been visiting lately – like Fairfax.He’s not doing these things for pointless reasons…. unless he’s become more unhinged than I’ve given him credit for 😉
Being a Qld’er, Dickson does seem to be inflated for the Coalition – both in an historical and demographic sense.
Sacha Says:
August 16th, 2007 at 10:38 pm
Is that 5 votes to 2 votes?
Yes.
Also 61% believe Costello said he wanted to destroy Howard.
39% said he didn’t.
One problem with multiple voting on NineMSN, it won’t let you vote more than once.
Unless there is some way around it.
Scorpio Says:
August 16th, 2007 at 10:57 pm
One problem with multiple voting on NineMSN, it won’t let you vote more than once.
Unless there is some way around it.
Change IP adresses, easy way on ADSL with no static IP is to disconnect/reconnect to get new adress, also works on dialup, but the latter is more expensive, or you could go via an anonymous proxy server which hides your IP addy as well.
Its the same method by which South Australia won the premium spots on the australian monopoly board this year !
I hesitate to change the direction of this thread, (it has taken a long time to get through all the posts after leaving it for a few hours) but there are some thoughts on the Queensland situation I thought are worth adding.
Ther have been a number of comments that a challenge would not be heard in the High Court until after an election is held and therefore the law would stand. However if you look at how the Bill is constructed it says a State or Territory law is of no effect etc. It seems to me that this does not stop the Queensland Govt from seeking to apply their law. It would be up to the Commonwealth or the Queensland Nats to challenge that law in the courts and have the courts set it aside.
I also wondered what the Howard Govt response would be if there were a plethora of requests from councils and others for plebiscites on a range of other issues. There is a limit to the capacity of the Electoral Commision to run these things- in fact I heard that the AEC was sh*tting itself at the prospects of the Queensland plebiscites. On what basis do you defend one set of plebiscites over another, particularly if they were for important issues like locating a nuclear power station in the council area.
Thats alright, give the city of churches something to crow about.
Rudd seemed to come a bit out of his shell on Lateline tonight. After failing to take the bait on the NT intervention, Haneef, Q’ld council, tassie hospitals … (on it goes) – He made an emphatic point on selling uranium to India – he’ll reverse the decision. Has he taken up the right issue?
I noted he used the phrase Howard is making up policy for “next ten weeks rather than the next ten years”. I reckon we might hear that line a few times more.
Possum, using census data to analyse the variation in swing at an election that has already occurred has not always been very fruitful, so I’m really very doubtful about reversing the process and using census data to predict the variation in swing at an election that has still to occur. I’d point you to an article using census and survey data published in the Australian Journal of Political Science showing that interest rates were not a hugely important at the 2004 election. I know a few campaign professionals guffaw at a conclusion like that.
Beware of data mining. The first use of the census data at British elections in 1970 found the two variables that had the strongest relationship with voting patterns was access to a private bathroom and access to an indoor toilet. After that, in the late 1970s everyone switched to factor analysis to deal with collinearity, but I note trend is back to using vast numbers of variables. You chuck enough data at a model, eventually you’ll explain a lot of the variability, but at the expense of a real understanding of what’s going on.
Paterson and Dobell are two of those seats that I believe are on an inflated margins. They both swung 5.5% away from the ALP in 2004. I live just outside of Paterson, in Hunter and some of the swings to the Libs in some of the strongest ALP booths such as Tarro and Beresfield were unbelievable.
Paterson will swing back and the redistribution will help a little. It is one of those bolters that Possum is talking about.
Why is Howard making so much noise about a load of insignifigant National Party hacks losing their council jobs? Because he’s in danger of losing a swag of QLD regional seats. Once again, it’s all about the next election and his own political survival that motivates the Rodent: very transparent.
Ben, in 2004, the Coalition flooded regional television with ads and Labor did very little. You couldn’t get away from Bob Baldwin on Newcastle television in 2004. I don’t think the Liberal Party will have regional television to itself in 2007.