If it wasn’t for the events of the last few days, today’s Newspoll wouldn’t be particularly interesting. It confirms the previous four polls, starting with December; Labor has reversed the slight dip it suffered in the last poll, and in fact extended its lead slightly, to 57-43 two-party-preferred.
That’s a swing of just under 10%, which, if realised on a uniform basis, would win Labor a 56-seat majority. Sounds far-fetched, but it’s entirely in line with the most recent results from ACNielsen (58-42) and Gary Morgan (57-43).
The interest comes from the recent somewhat fevered commentary to the effect that the Brian Burke affair had ended Kevin Rudd’s honeymoon period. If so, it’s not showing up yet in voting intention. But even that’s probably not very informative, since the poll was only taken at the weekend, and voters take a while to digest things.
Paul Kelly on Sunday’s Insiders was cautious, saying “I’m not convinced that it necessarily will end the Rudd honeymoon – we’ll have to wait and see on that”.
That prompted one blogger yesterday to ask: “What do you know that we (and your colleagues) don’t? Peeking at the Newspoll numbers again?”
Looking away from the figures for voting intention, however – something I don’t normally encourage – Rudd might seem to have come off the boil a little. His approval rating is down from 68% to 62%, and his “beauty contest” lead as preferred prime minister has dipped by three points, now 45% to 38%.
This will give comfort to those on the government side who believe the theory – so often used in the attack on Kim Beazley – that a lead in voting intention is unsustainable when accompanied by low approval ratings.
But before they get too comfortable, they should re-read this piece by Peter Brent from last October’s Fin Review. Brent has analysed the Newspoll figures going back two decades and found that they actually support the opposite theory. Opposition leaders with low approval ratings tend to do better at the election than the polls indicate; those with higher approval ratings tend to do worse.
In Brent’s words:
If the Newspoll data points anywhere, it is in a surprising direction, suggesting that survey voting intentions are more sustainable if accompanied by low – rather than high – satisfaction ratings. It might be that high approval can artificially boost voting intentions.
The next few months could present a very interesting test of this theory.