The polls were a disaster in this election, the worst I have seen in Australia and worse then anything that occurred in America in 2016. The last polls had the Coalition primary vote at 38, Labor at 37, the result so far with 76% in is the Coalition at 41, Labor at 34, a 7% difference. So you have a 6% miss nationally in terms of the primary vote differential, far worse then anything I have seen in other countries. In QLD the miss was even bigger, the primary vote in the polls was LNP 36 to Labor 32, the actual result so far is LNP 43 to Labor 27, so instead of a 4% difference, we have a 16% difference, a 12% miss in the primary vote, I am pretty sure that has to be the biggest failure of polling almost anywhere in the past decade.
I am going to use the American methodology of comparing polling error not the Australian one which looks at swings based off vote share.
Nationally the polls on average had Labor up 52-48 on the 2 party vote, so far Coalition leads 51-49, so a 2% Coalition lead instead of a 4% labor lead, a 6% miss nationally, in QLD it was supposed to be a 51-49 LNP lead, actual figure is 57-43, a 12% miss.
The polling error in Australia makes what happened in the midwest in 2016 look small by comparison.
In terms of the actual results, I would say the most interesting result were the swings to labor in affluent northern Sydney and the swings to the Coalition in the more diverse suburbs of Western Sydney. In the northern Sydney seats, you had swings of 3.4%, 4.5% and 4.7% towards Labor in Bennelong, Bradfield and North Sydney, all northern Sydney seats.
Whereas in western Sydney you had swings of 3.9%, 6.4% and 5.2% towards the Liberals in Blaxland, Chiefley and McMahon, all western Sydney seats. Maybe in 20 years time, Liberals will be more dominant in western Sydney and labor in northern Sydney, who knows.
