Oz Politics Blog

According to Morgan, in mid August, primary support for the Coalition fell 2.5 per cent to 39 per cent. Support for Labor fell slightly, down half a per cent to 41.5 per cent. If preferences were allocated as they were at the 2004 Federal election the ‘two-party’ preferred vote would be Labor on 53.5 per cent and the Coalition on 46.5 per cent. Had an election been held in mid August, Morgan predicted Labor would have won easily with minor party preferences.

However, I am always suspicious of polls that suggest large movements over a fortnight when it is difficult to point to a plausible reason for the fall. Morgan’s explanation is:

“Primary support for the Coalition fell 2.5% to 39% (7.4% below their result from the October 2004 election) due to the standoff in the Senate between the Liberal Party and the Queensland Nationals over the sale of Telstra and Voluntary Student Unionism. The Labor Party should be concerned by their inability to capitalise on the situation and increase their primary support. However if an election had been held in mid-August the ALP would have won on minor party preferences.”

I am not convinced. I suspect this result is a bit of an outlier; little more than stochastic noise. Though to be fair, the odds at Centrebet have drifted a touch in Labor’s direction since the first of August.

Also of note, we are starting to see the same persistent pro-Labor bias in Morgan (alternatively, a pro-Coalition bias in Newspoll) that we saw in 2002 and 2003. This time it is even more difficult to explain as both Morgan and Newspoll are allocating their preferences with a similar methodology.

The usual graphs are here (you may need to hit the refresh/reload button on your browser to see the latest).