I am hoping to have a Senate prospects piece up by the end of tomorrow but before that I think I need to address something I am hearing too much of at this election. I refer to trendy talk about federal elections being seat-by-seat battles, the election being 151 separate contests, and the idea that even a massive 2PP win (should Labor be so lucky) might seriously not result in a Labor majority. We hear this stuff at every federal election but this time around the commentariat seems especially awash with it. This piece is pretty mathsy (let's say 3/5 on the Wonk Factor scale) but it's necessary to crunch a few numbers to show what a load of nonsense we are hearing.
The state of play
Firstly a brief word about polling since last week's roundup. Two polls have been released, a Newspoll with Labor 53-47 ahead (L-NP 36 ALP 37 Green 10 ON 4 UAP 3 others 10) and a Morgan with Labor 57-43 ahead by Morgan's respondent preferences (L-NP 32.5 ALP 36 Green 12.5 ON 5 UAP 1.5 IND 8.5 others 4) . However, if last-election preferences are used the Morgan only comes out at 55.4 to Labor, so Morgan's headline 2PP not changing can be ignored. (The expected value for the Newspoll off the published primaries is 52.8, but Newspoll would have a more accurate number based on rounding). By last-election preferences the average of the most recent polls from each major pollster is now 54.2 to Labor. There is a reasonably strong case based on the history of Australian polling generally that this is probably skewing to Labor a little bit and the real underlying voting intention is more likely somewhere in the 53s, but always beware of Nate's 1st Rule. In any case, history suggests that whatever lead Labor now has will probably (but not necessarily) reduce by the time of the final result.