slouching towards Bethlehem

This morning's AM described the Coalition's best case scenario:

CHRIS UHLMANN: Well Tony, that the most likely outcome is a Labor victory. But - and it is a fairly significant "but" - that a Coalition win is not out of the question, and it certainly hasn't given up.

Now, that's based on them sandbagging all of their marginal seats and making it difficult for Labor to take some or all of the 16 that it needs and having to go higher up the tree, if you like, for the low-hanging fruit that's available.

And the way that that would work would be Labor … sorry, the Liberal Party holding the line in Victoria. Now, they need five per cent to take Deakin there, and the seats above it.

And it appears that Labor has stalled in Victoria at around about 4.5 per cent of the vote. So, if that happens and they can just lose three seats in South Australia, two in Tasmania, maybe pick up one in Western Australia, hold the line in the Northern Territory, then the Labor Party is forced to find all of the seats it needs, pretty much, to win government in New South Wales and Queensland.

So, it would be a difficult task, but as I said, that's the masking tape and bailing wire Coalition victory. It could come apart at the seams.

I applied these assumptions to the ABC's election calculator. The first figure shows the Coalition assumptions with no change in NSW or Queensland. I think the outcome is hopelessly optimistic because, out side Western Australia, the polls continue to show a decisive Labor lead. If I were drawing a true picture of what I think will happen outside New South Wales and Queensland I would give Labor 1 or 2 more seats, with the possibility of a major swing in South Australia throwing the calculation completely.

The Coalition battle plan has Labor searching for 13 seats in New South Wales and Queensland. I assigned 8 seats to NSW and 5 to Qld and then adjusted the election sliders to find what swings would bee needed in those states to produce a Labor victory f 76 seats. That gave me my second figure.

I find it hard to see how Labor can win less than 76 seats. Within the 2 states I think Labor will do well in aspirational seats on the edge of Sydney where both Work Choices and interest rates are biting hard. I'm told Labor feels very good about North Queensland where there are 2 marginal seats to be had, and Brisbane where Kevin Rudd is the favourite son. I'm also told Labor is very encouraged by polling in Bennelong and North Sydney. In 1983 many seats went Labor in the bush because it's hard for rural voters to support government candidates when their towns are collapsing in the face of prolonged drought and where I suspect climate change is suddenly a very live issue indeed. Last year I travelled around rural NSW for a week and it was hard to get anyone to talk about anything but the drought and climate change in pubs from Jindabyne to Dubbo.

If the swings in SA, NSW and Qld go far above the numbers in figure 2 then we are looking at an earthquake, not a landslide.