- ReachTEL poll on 23 June came in at 51-49 in the Coalition's favour. This time the self allocated preferences largely reflect the flow of preferences at the last election.
- Galaxy poll came in at 50-50
Across the four models I run, here are the outcomes ... First the two-party preferred (TPP) model that assumes the house biases across pollsters cancel each other out.
The second model, also a TPP model, is anchored to the 2013 election outcome.
The third model is based on primary votes, anchored to the 2013 election outcome.
The final model is also based on primary votes. Like the first model above, it assumes that the house effects of pollsters cancel each other out (ie. they sum to zero).
Collectively, these models follow.