The Poll Bludger

Malcolm Mackerras emerged on Monday, as he always does at times like these, to deliver his curmudgeonly assessments of the state of play in The Australian and the Financial Review (subscriber only, but there’s no telling what might happen if you click here). Typically he has no more than this to say on the House of Representatives: "I have no idea how the Australian people will divide their votes and nor does anyone else. However, this ignorance does not lead me to predict a close result. In truth, a close result is merely one of the possibilities. It is just as likely to be a landslide one way or the other". On the subject of the Senate he is rather more forthcoming, predicting that only a bad turn for the Coalition will prevent them achieving their "magic 38" blocking majority in the Senate.

This seems a reasonable assessment. Usually the six seats up for grabs in each state will split evenly between the right and left, and while one of the left seats will usually go to the Greens or Democrats, the Coalition have few natural predators on their side. Therefore, the Coalition need only turn in two reasonably good performances in succession, which they have not achieved during the current government due to a very modest performance in 1998 when they were lucky to win their House majority. In the Senate they won only two seats in New South Wales, where the result was Labor three, Democrats one and Liberal two; in Tasmania, where one went to independent Brian Harradine; and in Queensland, where One Nation edged out National Party Senator Bill O’Chee.

Fred Nile, Hetty Johnston and Family First arguably represent threats to third place-holders on Coalition tickets in New South Wales, Queensland and elsewhere, but by and large Harradines and Hansons are thinner on the ground this time. If the Coalition is to be denied its 38 seats, it will more likely be the 1998 New South Wales scenario where it is they rather than Labor that drops a seat to the Greens or Democrats. For that to occur the Coalition vote in the relevant state would need to fall substantially below 40 per cent, in which case the Government will be in trouble.

Various happenings around the place have prompted the first set of campaign updates for the House of Representatives election guide. To save you the effort:

Kalgoorlie (WA, Liberal 4.4%): State Labor MP Tom Stephens accused the Liberals of working to have transients and Aboriginals removed from the electoral roll by lodging objections against the enrolments of voters when campaign material failed to reach the addressee. If that is indeed their aim it appears it may not be successful, an AEC spokesman telling The West Australian it was "difficult" to remove people from the roll as the Commission had to "send three letters and have them returned with information that the elector was no longer at that address before they could be struck off". It is certainly to be hoped so, because a narrow defeat for Labor in Kalgoorlie could mean a narrow defeat for Labor in the election, in which case the controversy would put the popular Bush-stole-Florida theory in the shade.

Adelaide (SA, Liberal 0.6%): Liberal member Trish Worth, who at times had presented herself as sympathetic to asylum seekers, made headlines yesterday when audio emerged of her telling a Justice for Refugees debate that detention could not be ended "like that" because when "you bring a dog into this country, or a cat from some countries … there are certain tests to be carried out". This would probably not have hurt her if her seat was in the suburbs or the regions, but in central Adelaide, her traditionally more sympathetic image would have been the better bet electorally.

Lyne (NSW, National 11.3%): Rob Oakeshott, the independent member for the state seat of Port Macquarie, announced yesterday that he would not run against deputy National Party leader Mark Vaile, as he had been threatening. Given that Oakeshott scored 70 per cent of the primary vote at the 2003 state election against a Nationals candidate who received strong backing from Vaile, this would be a great relief to him and his party. This deprives the election of its best chance for a new independent to join Katter, Andren and Windsor.

Kennedy (Qld, Independent 8.3%): Bob Katter’s step-mother Joy Katter publicly endorsed his National Party opponent, James Doyle. The second wife of Bob Katter Sr, who held the seat from 1966 to 1990, Joy Katter said yesterday that she did not approve of his performance since leaving the National Party and that his late father wouldn’t have either.

Swan (WA, Labor 2.1%): Kim Wilkie has kept his Liberal opponent Andrew Murfin squirming over the letters scandal as the campaign begins, lodging a complaint with the Police Commissioner over the matter.

Election watchers with access to Sky News are advised to make a date at 9:30pm next Monday for The Gallery, a campaign chat show with a panel consisting of press gallery denizens Malcolm Farr, Glenn Milne, David Speers and Helen McCabe. Among the guests on last night’s debut was Sol Lebovic of Newspoll, who discussed the results of the weekend’s survey three hours before they became available through The Australian. In light of the Poll Bludger’s expressed doubts over the accuracy of last fortnight’s poll, his analysis included a number of well-made points:

I think that the poll two weeks earlier was actually a reflection of the previous fortnight of that particular poll which was all about the FTA … I think the whole Scrafton thing suggests that people who worry about that kind of thing and accept those comments about John Howard have already parked themselves on the other side … I think there’s a lot of voters out there who are sort of swaying in the breeze and the FTA got a lot of good media publicity for Mark Latham for those two weeks, we did a poll on that particular weekened and they said, "yeah, Mark Latham’s okay", two weeks later they’ve tended to forget about the FTA and we’re back to where we were all through June and July … the onus is very much on Labor to prove that they’re going to be better (than the Coalition). I don’t think we’ve seen enough out of Labor and Mark Latham in terms of the policies, they say they’re going to release it during the campaign, their tax policy for example, that’s a pretty big ask because that’s assuming they get clean air time during the campaign, and I suspect that voters in the early part of the campaign aren’t paying that much attention, they’re just picking up generalities. We know from polls we’ve done after the last three federal elections that a quarter of voters say they didn’t make up their minds until the final week of the campaign, so I think Labor’s got a long way to go.

In an election campaign in which both sides have nominated trust and honesty as the major themes, the Poll Bludger got off to a poor start yesterday with his assertion that he had "meticulously sifted through his House of Representatives election guide" in producing his Election Projection outcome (see left). In fact this was done in such haste that the Victorian seat of La Trobe was lazily marked down as "Liberal retain" despite a long-harboured expectation to the contrary, owing to the loss of retiring Liberal incumbent Bob Charles’ personal vote and the high profile of Labor candidate Susan Davies, who held the state seat of Gippsland West as an independent from 1997 to 2002 (many thanks to reader Barney Langford for pointing this out). Thus does the Election Projection now record Labor at the all-important 76-seat mark at which they can form a majority government, however narrowly. Hopefully this stunning breakthrough will give the Opposition Leader the shot in the arm he needs following this morning’s disappointing Newspoll result.

Another dubious aspect of yesterday’s entry was an implicit assumption that all three existing independents would line up with the Coalition if push came to shove. In fact the most likely outcome in the event that Labor wins 75 seats would be that Peter Andren, the independent member for Calare, would accept a Labor offer of the Speaker’s position, thereby giving them a majority on the floor. Andren’s seat was held by Labor before he won it in 1996 and unlike his cross-bench colleagues he has no history of involvement with the National Party.

The Coalition has bounced back from a horror result a fortnight ago in today’s Newspoll, taken over the weekend. Their primary vote is up from 39 to 43 per cent with Labor down from 42 to 40, though Labor maintain a 52-48 lead on two-party preferred (down from 54-46). Dennis Shanahan does a much better job than most journalists of talking up the poll’s significance without descending into hyperbole, saying the result "appears to bear out the Prime Minister’s view that the public is ‘fed up’ with the children overboard affair", which is perfectly true as far as it goes. However it is also true that the last poll probably inflated Labor’s lead.

The latest low-sample Taverner poll was published on Sunday, but the only information that may be gleaned from it comes from this article in Sunday’s Sun-Herald. Labor are on 53 per cent two-party preferred, showing no change on the poll from a fortnight ago, while the Greens are on a slightly implausible 11 per cent – also unchanged from last fortnight.

With three consecutive election wins under his belt including one that delivered him one of the biggest parliamentary majorities in Australian history, it would seem safe to assume that the Prime Minister is a reasonably effective election campaigner. But for political commentators being contrary goes with the territory, and the safer an assumption appears the more certain it is that a counter-argument will emerge. Peter Brent at Mumble has long been a Howard naysayer, citing the Government’s brush with death in 1998 and a victory in November 2001 that could be reckoned quite modest under the circumstances. The notion was given wider circulation today by Laura Tingle of the Australian Financial Review, who opened an article that featured prominently in today’s election wraparound thus:

John Howard’s utter dominance of federal politics for most of the past eight years has tended to overshadow something that will prove crucial to the outcome of the 2004 election: that he is a poor campaigner … (in 1996) Newspoll recorded the coalition’s primary vote falling from about 50 per cent in mid-January to 47 per cent on election day, March 2 … (in 1998) the government’s standing with voters deteriorated from a high of 44 per cent in mid-August to 39.5 per cent on election day on October 3. Even in 2001 in an election remembered for its focus on border security, the coalition’s primary vote fell from 50 per cent in late September to 43.1 per cent on polling day, November 10.

Tingle is being a little cute here. The 50 per cent result she chooses as her reference point for the 1996 election showed a spike in the Coalition vote for that fortnight, comparing with 47 per cent in the previous poll and 48 per cent subsequently. These results indicate that Howard maintained the status quo during the 1996 campaign, which would normally be considered a good effort from an Opposition Leader facing an entrenched and (normally) skilful incumbent. The poll chosen to represent the state of play going into the 1998 election was quirkier still, the quoted figure of 44 per cent comparing with 39 per cent from the previous fortnight and 40 per cent from the next. Following the logic of her 1996 comparison Tingle should have gone with the latter poll, that being the one taken immediately before the election announcement, but this would have required conceding that Howard held his ground during the campaign period. And if Tingle wanted to gauge the effect of a "focus on border security" in 2001 she could hardly have chosen a worse starting point than the poll taken immediately after September 11. It was inevitable that the Coalition vote would cool a little from this result, although the 43.1 per cent they ultimately recorded indeed represented a drop on their scores in Newspoll throughout the campaign – although it could just be that the maligned Roy Morgan was recording Coalition support more accurately.

The Government indeed has the fight of its life on its hands, but it may at least be said that any Labor partisans who are factoring pro-Labor bonuses into current polling based on Howard’s supposed campaigning incompetence are likely to be disappointed. Due to the inexperience of the Opposition Leader, for which the only modern precedent is John Hewson in 1993, and the proven capacity of governments to squeeze out one last term through election period scare campaigns, the Poll Bludger will continue to lean the other way and assume that the Coalition will claw back at least 2 per cent on two-party preferred over the next six weeks.

UPDATE (31/8/04): This post originally contained the outrageous claim that Howard’s 1996 majority was the biggest in Australia’s history. Thanks to Peter Brent at Mumble for alerting me to the fact that I had "taken the Howard truth serum" – it was in fact equal third.

The Poll Bludger has meticulously sifted through his House of Representatives election guide to determine the likely outcome in each of the 150 seats, which will no doubt be revised many times between now and October 9. It is certainly to be hoped so, because the current assessment is for the nightmare scenario of a hung parliament in the true sense of the word, with the Labor and non-Labor forces gridlocked on 75 seats each.

In New South Wales, Dobell, Parramatta and Richmond are nominated as Labor gains. Despite margins of just 1.5 and 1.7 per cent, Paterson and Eden-Monaro are tipped to hold against the tide due to the advantages of incumbency and popular local members, although most would say these were bold calls at this stage. Labor will recover Cunningham, lost to the Greens at a 2002 by-election.

A status quo result is predicted for Victoria, except that Christian Zahra is tipped to hold McMillan which became notionally Liberal with a margin of 2.9 per cent in the recent redistribution. Here the advantages of incumbency will work in Labor’s favour, but in three seats on narrower margins I am tipping the Coalition to get over the line. The bravest of these judgements is Dunkley, an outer suburban mortgage belt seat where Coalition scaremongering over interest rates, combined with the local issue of the Bracks Government’s tolls for the Scoresby Freeway, will bite as the campaign develops.

Queensland is universally reckoned to be the danger zone for the Coalition, and so it appears on this assessment – Labor will even up the ledger in this state by winning Herbert, Hinkler, Moreton, Longman, Bowman, Petrie and Dickson.

Three seats are there to be won for Labor in South Australia – Adelaide, Hindmarsh and Makin – and the current tip is that all of them will be, the first two by natural correction, the second with help from a retiring Liberal member, and the second with help from a sitting Liberal member (Trish Draper).

Labor will win a third successive clean sweep of Tasmania, win both seats as usual in the Australian Capital Territory and gain the Darwin seat of Solomon.

So on the current hypothetical election night, that will leave all eyes on Western Australia which would need to deliver three seats to the Liberals late in the evening to keep John Howard in the Lodge – or two to confuse the situation. The latter seems the most likely outcome at this stage, a projected 2 per cent swing to the Coalition in Perth delivering them Stirling and Hasluck with Swan to remain out of reach due to a slightly bigger margin and a comical pre-election performance from Liberal candidate Andrew Murfin.

A bleary-eyed and hungover Poll Bludger awoke mid-afternoon to face the big news, that October 9 will indeed be the big day. The Senate however will sit as scheduled on Monday and Tuesday, the Prime Minister saying he "didn’t want anybody to suggest that I was trying to prevent the Senate doing any pointless political business that it might want to do". Time now to take a Tylenol and a step back – individual seat predictions and a survey of the battlefield will be added later this evening, or maybe tomorrow (more likely tomorrow).