Newspoll: ALP 56-LNP44 (or let’s see things that aren’t there)

The hard sell started early on this one.

Last night at midnight, via the Mumble website, I found that the Newspoll to be published in today’s Australian was to have a 2 Party Preferred result of 56-44, a change from the last one of 57-43. In effect this was “no change”, as a one percent shift is well within the margin of error.  I slept well, having been a bit worried that Abbott might give the Libs a boost – as often happens (Alexander Downer for example took the 2PP vote from 48% to 54%!).

And yet when I woke and was idly swallowing my Light and Tasty cereal, what did I hear on Sunrise? Apparently Tony Abbott had delivered a poll “bounce” to the Liberal Party and had caused “the Liberal Party vote to jump by one percent”.

I then got to work and had a look at The Australian's website. Here’s what its website front page looked like:

Abbott gamble pays off for Libs

Dennis Shanahan, Political editor LIBERAL Party support has bounced back and Tony Abbott has cut into Kevin Rudd's lead within a week of taking over as Liberal leader.

• PANIC: Rudd calls off early dash to Copenhagen
• HOSTILITIES: Role reversal as leaders line up
• BY-ELECTIONS: ALP voters turn to Libs
• MEDIA: Selling the Mad Monk
• DENNIS SHANAHAN: Voters switching back to Coalition
• GLENN MILNE: A sign that Libs have got it right
• DAVID BURCHELL: Abbott must seize moment, if he can
• EDITORIAL: Rot stops as Liberal base embraces Abbott

Wow! I do like some balanced reportage.

Here’s what Dennis Shanahan had to say:

LIBERAL Party support has bounced back and Tony Abbott has cut into Kevin Rudd's lead as preferred prime minister within a week of the newly elected Leader of the Opposition spectacularly reversing the Liberals' stand on climate change and rejecting Labor's ETS.
…
The Newspoll survey, conducted from Friday to Sunday, exclusively for The Australian, showed a rise of four percentage points in the Liberals' primary vote, taking the Coalition's support to 38 per cent compared with the government's unchanged 43 per cent.

He also wrote:

The Newspoll survey suggests that the Liberal Party members and voters who were moving away are now returning.

Sigh. Ok, let’s work this through. Yep the Liberal Party did see a rise of 4% from 30 to 34. The problem is the ALP didn’t change from 43%, which means Abbott didn’t take any votes from Labor. So where did they come from? Well the National vote dropped from 5% to 4% and “Others” (generally regarded as primarily right wing parties like One Nation and Family First) fell from 10% to 8%. So 3% of the 4% jump came from the right wing, which is why the 2PP only increased a trivial one percent. So well done Tony, you brought back votes to the Liberal Party that were already going to end up there (or is he worried that Family First will run candidates against them??!)

Now about that “bounce”. Possum over on Crikey brought the logic and reason back into the debate. He looked at the last seven opposition leaders (ignoring Nelson and Crean as they went backwards). He compared the first poll after their leadership began with the average of the previous six weeks (to get rid of any rogue polls, and also because generally when there is a leadership change it has occurred after a sustained run of bad polls). Here’s what he found:

As you can see, Abbott has had bugger all impact; certainly not anywhere near Rudd’s league, and in terms of impact on the Government, nothing at all like the affect Howard had on the ALP back in 1995.

It is also interesting to note how low the Liberal (Government) vote was prior to Rudd taking over from Beazley – 41%; the lowest at any such similar stage since 1994.

The other point to remember when you see Howard had the Liberals at a higher vote than what the ALP is currently, is to remember the ALP gets more preference votes than does the Liberal Party – mostly because the Greens primary is around 10-11% but the National Party’s is only 4-5%, which means to win an election the Liberal Party primary vote NEEDS to be above the ALP primary, and above it by a good 4-5%. 

Possum then looked at the preferred PM ratings of these opposition leaders:

What he found was that Rudd is still dopily popular (and more so than he was when Turnbull took over!), and also that only 17% of voters are undecided about Abbott. More voters have already made up their minds about Abbott than they had Turnbull, Rudd, Latham and even Howard. So there is very little room for growth in Abbott’s figures. He can’t count on winning too many undecided voters, because 15% of voters are always undecided (ie they don’t give a stuff!).

So all Abbott has to do is win over all those people who currently think Rudd is doing a good job (his satisfaction rating actually increased from 56% to 58%).

Now look, maybe he can do it. But the last two leaders to win an election off the incumbent were within 5% (Howard 95) and 3% (Rudd 06) of the PM when they took over – ie already within striking distance, especially when you consider the advantage the incumbent has in such a poll. Abbott is within 37% or Rudd!

For Abbott to get up to the 36% that Rudd was when he took over leadership of the ALP in 2006, Abbott would have to increase his appeal by over half – or 13% of the electorate need to change its mind. And even if all of those 13% came from Rudd, Rudd would still lead the preferred PM 47% to 36% – in other words Abbott would still trail by 13%!

So is Abbott’s task possible? Well yes I guess so, anything is possible. Has it ever happened before? Well… no.

So when The Australian (and the Liberal Party) wants to talk about Abbott making a move, and Rudd’s support crumbling, they might do better to take a deep breath, and repeat to themselves at 56-44 the ALP would win the election by on average 101 seats to 46.

That ain’t close. That’s actually annihilation.