“Top story in tonight’s news – generalissimo John Howard is still dead” might as well be the headline for the latest crop of stories around the recent Newspoll result showing that almost nothing had moved in any significant direction as regards Rudd vs the Rodent and Labor vs the Coalition, all holding steady at around 57-43.

The utter absence of anything resembling news in the results has forced the pundits to start examining the nature and conditions of the vacuum.

“These figures are counterintuitive”, writes Michelle Grattan, “there’s been a feeling in Canberra that Kevin Rudd has gone a little off the boil”. İn the SMH, Peter Hartcher tries to work through the meaning of wide differences between Rudd’s overall approval rating and his low rating on economic management etc etc, while in the Oz, Glenn Milne suggests Rudd may lose via an incredibly torturous comparison with the distant and very different 1969 election – only slightly undercut by the fact that Gough Whitlam took 32 seats from the Coalition in that poll, and Kevin Rudd needs, erm, 16.

Could we send the troops in to the political bureaux to stop this pointless, utterly meaningless non-commentary? İ can’t think of any other Western country in which poll tea-leaf readings have become such a substitute for analysis of actual policy and ideas. How is is that someone like Grattan can reach the point where a relatively solid expression of intent over time, undisturbed by minor non-scandals and spin, can seem counterintuitive?

Presumably because polling combines two core intellectual errors. The first is to implicitly apply behavioural psychology – and the notion of the reflex – to the public, rather than thinking of them as people who reflect, form world-views, actively form their opinions. The current mantra around polling is that you just have to hit a nerve – the Kirribilli nerve, the dodgy childhood nerve – and people will reflexively react in a measurable way. Hence, any response that indicates that people aren’t amoeba becomes “counterintuitive”.

The second error is to apply a physical science framework to matters human, and thus to come to the false conclusion that an “opinion” exists as a discrete object, like a planet orbiting a star, whose trajectory is mappable, independent of contact, question asked etc.

Given that people 1) think economic management and foreign affairs leadership are the two key priorities of government, 2) think Howard is better at these than Rudd and 3) prefer Rudd, it should be obvious that opinion can shift from question to question even at the individual ‘pollee’ level. And when an election is announced, things will change again.

All we know from these thousands of pages of analysis is that people seem to prefer Rudd and Labor quite significantly and consistently – and even that may be an imaginary relationship, as when the fantasy of quitting work and studying mime takes over to such a degree that you begin checking for cheap flights to Paris. The fantasy ends when the credit card bill comes in, and it may happen that Rudd’s magic numbers will turn out to be the same unreal reality. But there’s nothing in the numbers that will tell us that one way or the other before it happens.

So, in the meantime, team Canberra, please please write about something, anything, with a skerrick of content. If you remember how to.