Monday, July 1, 2013

Weekly aggregation: Labor in front

If you have been following my blog, you will know I have been exploring a before-and-after Bayesian model that allows me to quantify the size of the discontinuity in voting intention between the reigns of Julia Gillard and Kevin Rudd. I have called this the Rudd Resurrection Effect (RRE).

In addition to the polls I captured in the past few days, today (as promised) I have backed-out the snap Morgan SMS poll (I don't have sufficient data to calibrate the house effect). In its place I have dropped in the latest Morgan poll, which has Labor in front 51-49.  I have also added Newspoll, which had it the other way around.

I have decided to exclude today's Essential poll as the current polling period spans both Julia and Kevin. Exclusion was necessary to maintain the integrity of the before-and-after capacity of the model. This is a bit of pain, because I suspect I won't get a clean Kevin poll report from Essential for another two weeks.

The final aggregation has Labor just ahead (and a Kevin effect approaching 6 percentage points). For a number of reasons, I think we need a couple more weeks worth of data before we can be confident of the level of TPP support for Labor suggested by today's polls.

Elsewhere, Pollytics has an interesting post on the impact of uneven state swings. His conclusion is that a 5% swing to Labor in Queensland would deliver around 9 seats simply because so many Coalition held seats sit on relatively small margins. He argues that Queensland's innate support for Kevin Rudd could see him win a poll with a national TPP of 51-49 in the Coalition's favour.

No comments:

Post a Comment