Correction to weekly aggregation

Yesterday I excluded this week's Essential poll because I mistakenly thought the poll included some of Julia Gillard's prime ministership. A few people emailed me with the simple news: I was wrong. The one-week sample from Essential is solely from the second Rudd period.

So this is me eating humble pie. I have added in the weekly Essential report into the dataset and re-run the Bayesian model. Rather than Labor a touch ahead (on 50.1 per cent TPP as I reported erronously yesterday), the model now has the Coalition on 50.2 per cent TPP. Not a lot of difference.

The Rudd Resurrection Effect is looking like 5.6 per cent. Still substantial, and still a potential game changer.

A uniform national swing has us back at the hung Parliament. However, as we noted yesterday, the swing to Labor is unlikely to be uniform. Everyone expects a larger swing in Queensland where Coalition seats would fall in quick succession.