Over at The Oz there’s a pretty funny Editorial, where yet again the owners of Newspoll are getting all confused about the reality of their own product – seemingly now choosing to ignore everything it has, or has ever had, to say. The first line of the article sets the scene:

The next election is wide open and will be won or lost on a single issue — economic management credentials

Yeah, I know – the Op-Ed pages are turning into a fruit machine, where you whack in your coin, pull the handle and get the same old lemons every time.

As far as public opinion is concerned, “Economic Management” as an electability issue doesn’t behave as a continuous variable with transitive qualities – the party that scores 50% on “better economic manager” is not statistically more likely to win an election than a party that scores only 45% on the same question. The party with the best score may happen to win more elections than the party with the lower score on this vague concept, but there is no statistical evidence of causation – which can be clearly seen by looking at what happens when governments change at both the State and Federal levels. As some of you may remember, we went through all this in 2007 with a fine toothed comb and were shown to be correct, even though we didn’t own any pollster at all.

However, there is evidence to suggest that the better economic management question does behave as an ordinal variable on electability, where the order is defined by being above, at, or below some required community threshold of economic credibility – where as long as a party is perceived to be above that threshold and are seen as credible at managing the economy, seen to be trusted enough to manage the economy, the contest over economic management credentials ceases to be one defined by the importance of magnitudes.

A basic credibility on the economy, in one sense, can be seen to be a pre-requisite for an election victory as opposed to the size of the economic management gap between two parties being a deterministic indicator of an election victory.

The reason the Oz gets it wrong on this and most of the rest of its public opinion prognostications for that matter – yet only ever in the Op-Ed pages mind you – seems to come back to the inability of the Editorialists to come to grips with the actual political reality of the Australian population.

The article pushes the central theme that the ALP must move to the center to be successful. Well.. er, “Yeah” I here you say – and the sun rises in the East! Espousing ‘median voter theorem’ isn’t exactly profound.

But it’s where the Oz Central Command believes the center of Australian politics actually resides which is worth a giggle – they spend too much time looking in the mirror, mistaking their own reflection for political reality.

We actually know where the mean of Australian public opinion sits, as well as where the Australian public believes the major parties sit in terms of their left/right political perceptions and relative distance from the political center – we have the Australian Election Study to tell us. The two relevant questions in the survey are:

1. In politics, people sometimes talk about the ‘left’ and the ‘right’. Where would you place yourself on a scale from 0 to 10, where 0 means the left and 10 means the right?

2. Using the same scale, where would you place each of the Federal political parties?

We can tally the results for the 2007 Election survey up in a diagram and look where the actual centre of Australian politics resides, where the actual public perceptions of the major political parties reside and compare it to where Oz Central Command apparently reckons the public centre sits.


And they wonder why they get it wrong so often.

(Visited 26 times, 1 visits today)