Spending six months in a leaky boat (or is that vote?)

One of the rituals of election time is the leaking of internal polling. The major parties attempt to quantify what the electors tell their erstwhile campaign workers and weave it into a narrative. Any such polling should thus be regarded with industrial strength qualities of salt. What is interesting is the story the parties choose to tell by selectively releasing the results.

Labor's interest lies in painting the election as a foregone conclusion, while maintaining their underdog status. This is necessary purely to erode the collective groupthink that Howard is the greatest politician in history and he will hold back any wave of resentment with Canute-like resolve. As most published polls are doing the job for them, poll leaks have been rare except for the irrestible allure of claiming seats with a margin of 10% are in range. This blanket leak is designed to counter Howard's seat-by-seat strategy by portraying the swing as national, and particularly violent in traditional 'true liberal' strongholds, reinforced by a direct poll in North Sydney. They also nod in agreement whenever an obliging poll puts Howard behind in Bennelong.

The Liberals' strategy has come in three parts. The first part was the reactive phase, with Howard forced to ask retiring members not to retire after all with the seat of Grey being emblematic of his plight. At this time, Howard had not yet closed the window for a change in leadership and the government was observing a trend it simply couldn't comprehend.

The second phase was heralded by that familiar bellwether, Eden Monaro. As Howard stared down his gutless wonder frontbench, he revealed internal polling that said Mr Nairn was doing just fine and that Roy Morgan's 60-40 poll was a fantasy. This didn't seem to make a lot of sense given the Newspoll marginal seat average for July-September was 58-42 to Labor, remarkably similar to a WIN poll conducted in Eden Monaro. The symbolism of a favourable poll in the 'national bellwether' seat was palpable: we're just fine because Gary's just fine. Once Howard decided to stay, annihilation became business as usual in the course of one afternoon.

This departure from apparent logic had the effect of making Malcolm Turnbull look as traitorous as Rommell after the failed assassination of Hitler. For his sins, Malcolm has to cop a 98% disapproval of the Gunns Pulp Mill in his electorate.

Since Howard's evocation of Eden Monaro internal polling, it has popped up as a weapon in ideological struggle. When Labor trumpeted another poll in North Sydney, it was contemptuosly squashed, with the statement 'workchoices doesn't rate in this seat'. In Tasmania and Queensland, where polls consistently have the Liberals in dire straits (beset by Sultans of Swing from Devenport to Cairns), internal polling has backed Howard's national judgment. Exact polling figures were released for Lyons, which showed anti-mill (ex)Liberal, Ben Quin, suffering a 11% primary swing. The other polling for Tasmania was somewhat obscured by a News Ltd sieve, but the messenger Glenn Milne did not offer any numbers to back up the assertion that 'the Liberal candidates in all other seats in Tasmania (were) improving'. Given the Tasmanian local polls had Labor at 60% in Bass and Braddon, this seemed to require a certain suspension of belief. What appears to be happening is the Liberals are conceding they can't win Lyons off Labor. All the opposition is concentrated and caused by local issues. This thinking evokes the 16-seat-Everest image. If the Libs' polling was true, there would be no net seat loss and the Mersey expedition would be a success, thus vindicating Howard's marginal seat strategy and presenting the picture that jobs trump trees in voters' minds.

Meanwhile in Queensland, the Liberals have tried to paint a disaster-in-the-making in Ryan as collateral damage of a local pork-barrelling mission. Liberals have clearly decided that they can't cover up Ryan's problems. Michael Johnson, from all reports a very unpopular local member, has tried to blame his plight on the Goonda bypass, built to assist the seat of Blair. Johnson has been left to make desperate pleas for assistance from the raging Rudd hordes, reminiscent of the British epistles to Rome shortly before the Saxons rained down upon them.

This damage limitation has both a quantitative and qualitative effect. First, it means that bad national polls do not translate to inevitable defeat - meaning the polls must by simple mathematics narrow once the election is called. Second, it contradicts the idea that Howard's policies are the problem: either it is the candidate's bizarre flirtation with greenies or the politics of land resumption. The idea is to present Howard's policy as correct, the true liberals as obediently voting on the economy as per usual and the blue-collar 'battlers' are backing Howard's mantra of job creation.

The ALP has quite simply had enough of this posturing and released a fusillade of counter-polling. The talk is of 'vote north of 55%' in seats including Flynn, Herbert, Hinkler and even Leichhardt. Interestingly, either the Queensland Libs have gone native or they haven't started outright lying about their own polling, as they 'struggled to dispute' Labor's assertions.

Based on the above, brave predictions include Liberal polling to reveal a massive swing in Lingiari and Solomon remaining with the Libs in NT, Liberals set to gain Issacs and Holt on the back of anti-Sudanese feeling in Melbourne yet suffer no swing in Higgins and Libs to pick up Cowan while holding Stirling in WA. The possible problems in the seat of Canning will be blamed on the unpopular brickworks proposal.