One of the things which Graham mentions is the heavy investment in TV advertising by labor.
This medium is almost entirely based on image - it is used to establish perceptions about a product or issue, and then confirm them.
One of the things which Abbott and his cohort have working against them is their apparant arrogance, particularly on the part of Morisson, over a refusal to inform us, their masters, of what they are doing and why.
I suspect that they are falling victim to the increasing encroachment on information sharing by the "intelligence" communities.
In any case, Abbot's smug, patronising demeanour is a very bad component to flaunt in any image building campaign, and I suggest it gave great credibity to Labour in its "knocking copy".

Graham,

I would have thought that there was little left of KR's personal following after the 5.something swing against Labor in 2013 (he had apparently built it up in two previous elections). And while I see the argument for using Senate voting as an index of true party support, people can and do see the two contests as separate, and will vote accordingly.

I can't see any sign of a voter backlash in the by-election result, and I would think it a good outcome for the Government.

So how do you explain the difference between his first preference vote and the Senate ALP first preference vote Don? It is quite striking. I also used the same method to predict that Peter Slipper probably had a negative personal vote, which his meagre showing as an independent tended to bear out.

Dear Graham,

I sometimes think you are ever the contrarian?

Picking up this Griffith “dead rubber” for even more reinterpretation seems futile but it also sounds like a bit of fun, so here goes.

All sides of the result will tell us the same thing, The winner says it endorses their position and reflects badly on the loser. The looser says the winner actually lost primary support and claims a moral victory?

Nothing new here, unless you have something new? Like a statistical “model” perhaps?

I guess that your “modeled” interpretation will have to go through regular iterations between now and 2016. Having produced the GY1 from your model, could you do a “hockey stick” next?

I think I might just wait until 2016

I would agree with your general thoughts about Rudd' popularity/personal vote.

Perhaps Glasson benefited from the reverse donkey vote. Certainly the Stable Population Party picked up less than 1% from the top position. A breakdown of preference flows would be useful in a final analysis.

Your assert that "Labor's campaign was dishonest" but at least it focused on an issue (Medicare) that is important to voters. It wasn't just based on personal attacks, like the LNP video and other propaganda, portraying Butler as a union hack. Once upon a time it would have been called 'playing the man'. Sometimes good lawyers make good lawmakers but I can't remember the last former President of the AMA who did, despite it being Australia's most successful union.

I think you are missing one thing Graham. It is some time since I lived in the electorate, but I think it is still now the "public servant town" it was then. Not only that, it was a middle/top end public servant town.

The proximity to the city, & the property values have ensured this result. I was amazed at how many of my neighbors were bureaucrats, they were virtually wall to wall in my area.

I am amazed that the libs can do so well there, after the drubbing Newman gave the bureaucracy. They must be fuming still, that he could dump so much useless garbage, & the effect disappear without trace in a matter of weeks.

Newman set the stage for Abbott when he moved into Canberra, demonstrating just how over manned our bureaucracies are. Unfortunately Tony squibbed it, but it has still not attracted any love from public servants to the libs