The great campaigner?

Given Howard's perilous starting position in this election campaign, I've had a look at the Newspoll data for the last four elections and come to the conclusion that a Howard victory is now so remote it would be one of the biggest upsets in Australian history.

Using the primary vote as an indicator of solid support for the major parties, the history does not back up the notion of Howard as a great campaigner. He has only increased the Liberal/National vote in one campaign. In 2004, Howard added 1.7% to the Newspoll figures during the campaign and a further 1.7% at the ballot box.

Conversely, Labor's figures aren't great either, with ground being made up in Newspoll only to be lost at the ballot box. Latham is by far the worst offender, losing ground in both the polls (3%) and the final vote (1.4%) to go from competitive to thrashed in six weeks.

These figures indicate that Howard has made most of his gains prior to the election, with the electorate's flirtation with Latham seriously souring as the campaign went on. Interestingly, the electorate was volatile, with the Liberal vote bouncing around in a 7% range, while Labor fluctuated within a 3% range.

The really bad news for Howard comes when we compare the minimum/maximum poll figures for the 6 months leading up to the election. No party has scored a higher primary vote than it had reached in the previous 6 months, and Howard has the lowest maximum vote of the lot (41%).

The 6 month fluctuations demonstrate the polls to be as set in Rudd's favour as they were for Howard in 1996, with both parties' primary vote moving within a 6 point range (Labor 46-52, Lib/Nat 35-41), as compared to 4 points in 1996 (Lib/Nat 46-50, Labor 39-43).

A weighted re-run of 1996 would give Labor 47.5% of the primary vote and Lib/Nats a shade under 41%. This would give Labor a two-party preferred vote of somewhere around 54%.