Poll watch goes into full swing

IN Canberra everyone is keenly awaiting next week's Newspoll results.

Politicians tell the media they don't poll watch when determining policy and political strategy.

Journalists from rival media houses publicly spruik their own polling organisations, as they must.

But privately everyone knows Newspoll continues to dominate people's minds in terms of how both the parties and the party leaders are faring.

One of the reasons Newspoll is so dominant is because of its sample size and the frequency with which it comes out.

It is national, has one of the largest sample sizes going around - thereby helping to reduce the margin of error on its results - and, with very few exceptions, Newspoll is released every fortnight, meaning that trend lines are easily established.

When the coming week's poll was delayed from being released this week because of the Queen's Birthday long weekend, every politician noticed: that's how closely they hang on the results.

Former Liberal leader Malcolm Turnbull used to lament the fortnightly kick in the teeth he would receive from Newspoll, desperate for it not to include comparative leadership ratings, which showed alternatives such as Peter Costello and Joe Hockey were more popular than him.

The other reason Newspoll is so important is because internal party polling is not nearly as prevalent as people might think. Opinion polling is an expensive business and it involves a certain amount of navel-gazing. Parties often determine that even this close to an election campaign they are better off putting money aside for an expensive advertising campaign than polling for the sole purpose of knowing how they are travelling.

Knowing you are doing well or badly doesn't of itself help you do something about it - unless one is in denial, of course.

What parties do a lot of is qualitative research, such as focus group work.

It is cheaper and with the right collection of swinging voters from marginal seats, it helps them determine issues of interest and how public perceptions of the leaders are being formed.

Labor's national secretary Karl Bitar is a master of it, as was long time Coalition pollster Mark Textor. Where internal party polling does ramp up is closer to the election, in particular during the campaign.

That is when so-called track polling is conducted: small samples of voters, say 200, polled from one electorate (or a small geographic collection of key seats) at a time on consecutive days.

It builds a picture of how the public is reacting to the daily news cycle, and moving the poll around key seats helps the campaign organisers determine which seats need a cash or personnel injection from the central party to help them along.

Knowing how marginal seats are responding to leaders and policies is where elections are won and lost.

Getting back to next week's Newspoll, with the exception of surprises in the questions asked (you might think a question about preferred Labor leader between Kevin Rudd and Julia Gillard will be in the mix, but there could be more surprises), most commentators and strategists will be keeping their eyes out for the party primary votes and the net satisfaction ratings of the leaders.

The other two results - the two-party-preferred figures and the preferred prime minister ratings - are less important.

The best way of calculating two-party results is contested ground.

Some say using the last election preference flows is best; others point out those results never consistently hold.

Some say asking respondents the two-party question is important, but doing so carries a high statistical margin of error in the final results.

The preferred prime minister figures aren't so important because incumbents usuall dominate on that score, but that doesn't always mean they win the election.

What the primary vote says is important because it is pure, untainted by interpretation. It's the one-party figure that strategists closely watch. The net satisfaction rating is also important - the number of voters satisfied with a leader's performance minus those who are dissatisfied - because it largely takes the incumbency advantage out of the picture.

This close out from an election (most observers think next week will be the last parliamentary session before it is called), in normal circumstances neither leader should have to be too worried about the published polls other than for the picture they paint for the election contest.

But we are not in normal circumstances at the moment.

Rudd's support has fallen off a cliff and with the ongoing mining tax dispute many inside his party worry about whether he can recover.

Gillard has the numbers if she wanted to knock him off, but she appears to be prepared to risk Rudd at the election and only think about the leadership once the poll is out of the way (whatever happens).

As the third Liberal leader inside this term of government, Tony Abbott is safe, but Liberals have their attention fixed on the next Newspoll anyway because they don't want Rudd's support to fall away too much lest Labor loses its nerve and replaces him. Abbott fancies his chances against Rudd but the return of support Gillard as leader might generate for Labor's brand is more concerning.

What a strange situation - a Newspoll where everyone is eagerly waiting for the results, but they all want Rudd to do OK. We'll soon see if he does.

DURING the week I wrote a feature article about the role of three early 30s staffers in the Prime Minister's office - Alister Jordan, Andrew Charlton and Lachlan Harris.

It focused on their influence and their inexperience, highlighting the problems of a PM who doesn't listen to his cabinet or elder statesmen in his party but takes such close counsel from relative political and policy novices. I have never had so much feedback to an article from Labor insiders telling me I was dead on, and from members of the public concerned about the situation.

But we need to remember the key issue about this state of affairs is not any negative assessment of this trio. It is that a leader would be so weak as to surround himself with callow counsel. All political leaders past and present use (and need) youthful advice. And when those advisers are of a high quality, enthusiasm can sometimes trump inexperience. But wise heads are also needed, and in Rudd's case (unlike the cases of Bob Hawke and Paul Keating, who also had up-and-comers on their staffs) they aren't in the picture.

Some political insiders didn't like the fact the piece put staff in the public glare, suggesting they should remain in the shadows.

That's just ridiculous. Political staffers are an extension of their boss, which is why they can't be forced to appear before parliamentary committees. If the media doesn't scrutinise their role, no one will (or can).

There was one important error in the article, however, that should be corrected for the public record, although it in no way relates to the thesis of the piece. In setting the scene of Harris's role in the office I wrote that he came up with the Kevin07 slogan, which he did not.

It was in fact crafted by Neil Lawrence, the founder of Lawrence Creative Strategy. In the highly competitive word of PR, credit where credit is due.

Sorry about that.