Mark Latham is in front but how many political experts are prepared to call it with a few key campaigning days still to go.

From the second October 1 sealed section

So, how many of our senior journalists will conclude their big reads in the Saturday papers with the line that Prime Minister Mark Latham is now a very likely scenario? Crikey hears that Labor hard-heads are telling Iron Mark to shut up, stay calm and the keys to The Lodge are his because the trends in the marginals are heading in the right direction after he won the debate and then pulled out his Medicare Gold trump card.

There will be no triumphant, presumptuous pre-election saluting to the crowd a la the late British Labor leader Neil Kinnock whose premature celebration at the now notorius Sheffield Conference days before the 1992 British general election is widely attributed for his subsequent surprise defeat.

Check out this BBC link for the way Kinnock then blamed his loss on conservative newspaper editors and ask yourself whether the pro-Howard Murdoch press might fire up in the last week to simultaneously paint Labor and Latham black and favourites?

Afterall, then Daily Telegraph editor Col Allan gloated to Crikey on the Monday after the 1998 election that John Howard had rung him on the Sunday to say that his election day splash “Labor set to seize power” had helped him retain a couple of key NSW seats. Col said he used the word “seize” to deliberately create the impression Labor was getting something it didn’t deserve.

Still, with eight days to go till the polls close, there is still time for the PM to throw more money at key constituencies and today he was shovelling more bucketloads in the direction of the wrinklies.

That said, let’s now dissect exactly what sort of swing is on. As mentioned last week, the most valuable poll tool of the campaign is the weekly state by state breakdown of the Newspoll figures. Uniform national swings are rare, but state swings are often close to the mark for marginal seats.

However – two warnings – one, this is based entirely on Newspoll, which is showing a much better vote for Labor than Nielsen or Morgan. If they are right and Newspoll is wrong, then this exercise is meaningless. Two, as you will see, this election is very tight, and even a few votes either way could change this result.

BUT – if the statewide swings recorded by Newspoll are applied to the electoral pendulum at mumble.com.au then the following results occur:

WA – Labor’s vote is down 0.4 – this would produce no change, but Hasluck (1.8) and Stirling (1.6) are obviously in danger for Labor if this swing increases.

SA – Labor’s vote is up 1.1 – this would just deliver Adelaide (0.9) and Hindmarsh (1.0) but even the tiniest correction here would keep these seats in Coalition hands. Both these potential gains must now be considered very doubtful for Labor.

VIC – Labor’s already high vote is up 1.9 – this would deliver Deakin.

NSW – Labor up 3 – a big swing back that would deliver Dobell (0.4), Parramatta (1.2), Paterson (1.4), Richmond (1.7), Eden-Monaro (1.7) and Page (2.8).

QLD – Labor is up 8 – this would deliver a bagful of seats including Herbert (1.5), Hinkler (2.2), Longman (2.5), Moreton (2.6), Bowman (3.1), Petrie (3.5), Dickson (6.0), Leichhardt (6.4), Blair (6.6), Forde (7.0).

There is no state breakdown for Tasmania, where Labor holds all 5 seats and the only seat in play is Bass, or the ACT, where Labor will easily retain both seats, or the Northern Territory, where, on any swing, Labor is a strong show of taking Solomon (0.1) the most marginal seat in the country.

Assuming all the above, and that Labor regains Cunningham, the Parliament would look like this:

Labor 84
Coalition 63
Independents 3 (Katter, Windsor, Andren)

A very comfortable Labor majority of 18.

Assuming Labor does not achieve some of the more extreme results above (Petrie, Dickson, Leichhardt, Blair, Forde and Page) and fails in Adelaide (Adelaide and Hindmarsh), then Labor would have 76 seats – the absolute minimum for majority government and victory would depend on Labor not going backwards in Tasmania or WA and whether Malcolm Turnbull hangs on in Wentworth!

Newspoll (not Shanahan) says Labor to win 16-seat majority

From the second September 30 sealed section

By Crikey psephologist Charles Richardson

Today’s “Campaign Barometer” in The Australian, with accompanying story by Crikey’s favorite, Dennis Shanahan, could have been headed “Labor to win 16-seat majority.”

Of course, it isn’t. But that’s what the numbers say: applying the state-by-state 2-party-preferred figures to the pendulum (Tasmania and the territories not included) gives a Labor gain of 19 seats. Shanahan does say that “Labor has forged ahead in … Queensland, where it needs to win five seats to ensure a victory…”

Yet Newspoll’s figures actually show Labor would win 10 seats in Queensland, together with 6 in NSW, 2 in SA and 1 in Victoria. It’s a good illustration of how the pendulum works, too. The overall swing predicted is 3%, which on a uniform swing would be a gain of 18 seats. State by state, the detailed picture is quite different, but the total is almost exactly the same: the deviations from uniformity cancel out. (I’ll bet Malcolm Mackerras wishes he had a dollar for every time he’s had to say that.)

No-one really believes Labor will get an 8% swing in Queensland (which still wouldn’t quite be enought to put Dr Ivan Molloy into parliament). But if a few of the Queensland 10 don’t come off, there are plenty of other prospects in Victoria and South Australia. It’s swings and roundabouts: if the votes are there, the seats will follow.

Are the votes there? That’s what we’ll find out next week, but Crikey’s political editor Christian Kerr was openly tipping a Labor victory on ABC Victoria this morning.