Possum: how the once-mighty ALP fell

One of the more astonishing things about the federal election result is how the ALP managed to destroy such an enormous amount of public goodwill over such a relatively small time frame.

To really highlight the drama of it all, it’s worth looking at a couple of charts. Firstly, let’s take every two party preferred poll result between January 2008 and the 2010 election, and let’s also add a distribution curve of those results to the left hand axis. Secondly, we’ll look at the polling trend over the period with its uncertainty band (the all pollster, pooled sample trend plus the estimated margin of error band around the trend line).

What we see is a term clearly split into two distinct episodes — the period up to mid-September 2009, and the final week of September through to the election. The former represented a political juggernaut, the latter was something between a complete fiasco and utter disaster.

Something worth looking at is the way the satisfaction dynamics of the leadership of the two parties played out over the first term. What we usually see with these things is that there’s a relationship between a party’s vote and it’s leader’s satisfaction ratings — but where the relationship is much, much stronger for the party in government than for the party in opposition.

Labor’s first term was no exception — it was particularly orthodox in this respect. We can see that by running a simple scatter plot and regression line of the two party preferred vote for each side of politics against their leader’s respective satisfaction rating scores from Newspoll.

The Labor chart is much tighter and the regression line much steeper than the Coalition’s, showing us that the dynamics between the perception of the PM and the public support for their party is much more intimate that that of the Opposition.

So it’s worth charting the PM satisfaction with our ALP two party trend estimates. The two party preferred is on the left hand axis, the PM satisfaction on the right. 

What was interesting with this dynamic is how the vote estimates often turned a polling cycle or two before we witnessed any comparable change in the satisfaction ratings of the PM. Yet there were two noticeable occasions where that didn’t hold, and there was a shared commonality between these two events.

The first was when the Godwin Grech saga occurred, where the PM satisfaction rating started increasing a polling cycle or two before the vote estimates  started increasing. The other was in the election campaign, where Gillard’s satisfaction rating started falling a few days before the vote decline was picked up by all the pollsters.