All the sidebar trends and info have been updated for the latest polling cycle. Looking at the longer term picture, our Pollytrend since 2008 comes in like this:
The all pollster trend has broken away from the phone pollster trend over the last few weeks – for the first time on a couple of months – suggesting that the non-phone pollsters are picking up a slightly higher level of ALP support than the phone pollsters. On the primary vote trends for the majors, we get (click to expand):
As you can see, the turnaround in the Labor primary vote is much larger than the recent change in the Coalition primary vote, suggesting that there was quite a lot of Labor/minor party vote substitution going on of late as well as a bit of Labor/Coalition vote substitution. We can also look at a trend of the Greens primary vote and trend this year (click to expand):
It’s been an interesting little fall and recovery for the Greens over the last month, suggesting that some initial Greens voters moved to Gillard immediately after her ascension, but then some voters moved towards the Greens over the following weeks – whether they were the same group or not is a particularly interesting question. Nielsen’s respondent preference allocation data suggests Green to ALP preference flows have jumped back up to the level we’ve seen over the last few elections – approximately 80%. Considering it was down to as low as 68% at the nadir of the ALP primary vote, if it was mostly the same group of Greens voters who moved to Gillard and then quickly back again, Gillard’s strength has been bringing a few Coalition voters into the Labor column and realigning Greens preference flows to their historical norm.
If it wasn’t mostly the same group of voters that moved from the Greens to Labor and then back again, then we have some rather complicated compositional issues at work here with, maybe, one lot of ex-Greens voters moving to Labor after the Gillard ascension on the one hand, and then being replaced by pro-Rudd ex-Labor voters moving to the Greens and delivering preference flows back to Labor strongly on the other hand.
If the latter was the case, we would probably expect to see it manifest in Qld the strongest – and small sample Nielsen data did show that the Greens vote increased in Qld from 6 to 13 over the June 25 to July10 period. However, the Qld cross-tabs had MoEs attached of around 7%, so the evidence is only weak at the moment.
Just food for thought.