Monday, 17 November 2014

Chunky swing brings ALP back to post-Budget level

This week's Morgan and Newspoll must be somewhat concerning for the Government, with both reporting a two-party preferred of 55/45 in favour of the ALP. Here's how that looks in context:


That's an estimated 54½ as of last weekend, with a 90% confidence interval of about 1½ percentage points on either side.

I'm calculating the two-party preferred trend by using last election's preference flows applied to underlying primary votes. It seems that a few percentage points' worth of voters have switched from the Coalition to the ALP in the last week:


Translated into seat-by-seat outcomes, that would deliver a large ALP majority, with 87 seats won by the Opposition in the median case.


Monday, 3 November 2014

Two party preferred moves to 52/48

I've updated the model with this week's polls: Morgan, Newspoll, the new Ipsos, and Essential. Together, they've implied a small tick up in the ALP's primary vote, a continued decline for PUP, and basically unchanged for the Coalition and the Greens.


Using 2013's preference flows, that implies a two-party preferred of 52/48 in favour of the Opposition. Last week's movement towards the Government has largely been revised away in light of the more recent data.


The model keeps track of primary votes for the five parties (ALP, Coalition, Greens, PUP and Other) in each state and territory. Feeding those primary vote estimates into seat-by-seat simulations, using preference flows and residuals from 2013, you end up with a expected outcome of 76 seats for Labor; 70 to the Coalition; the Greens with a 2 in 3 chance of retaining Melbourne; plus Katter keeping Kennedy and Wilkie returned in Denison. The model's estimates of the last three seats, plus Indi, should obviously be taken with a grain of salt.

Since last week, I've updated the plotting code. The graphs now make it a bit clearer that the model's trendlines are fitted at a weekly frequency. Also, I've updated the election simulations so that they use a resampled set of 2013 residuals. You can see all the current data and code at the model's GitHub repo.



Tuesday, 28 October 2014

National 2PP trend: 51.5 / 49.5 to the ALP

I've added some code to the repo that will calculate a two-party preferred estimate from the new model, using last election's statewide preference flows.

The current estimate of the national 2PP trend is 51.5 / 48.5 to the ALP:


This two party preferred trend is based on the State- and Territory-based primary vote estimates that are at the core of the model. The dots are pollsters' estimates of the national two-party preferred. (I've added them to the repo too, so you can replicate the graph yourself if you like.) The shaded area shows a 90% confidence interval. The most recent polls on the graph are a few days ahead of the last displayed point of the trend estimate, which is calculated weekly.