Oz Politics Blog

I thought I would have a go at simplifying the blog presentation. I have rewritten the presentational theme files from scratch. It looks good on my Linux set up, under each of the browsers I have thrown at it.

Unfortunately, if you are still using Windo$e (or at least an older version like my family), the right-hand sidebar does not look as intended. If you want to see what you are missing, click on this link. Update: I think I have found a fix for Internet Explorer version 6.

Comments and suggestions welcome.

During February 2007, almost 8000 people were polled across three Morgan polls, two Newspolls and one ACNielsen poll. The headline prediction was a national two-party preferred vote share for Labor of 57.2 per cent, leaving 42.8 per cent for the Coalition. On the primary votes, Labor got a predicted 47.6 per cent. The Coalition’s primary vote prediction was 37.4 per cent.

The usual opinion poll graphs are here. You may need to hit the refresh or reload button on your browser to see the latest graphs.

Morgan has released its past fortnight polling as two results, one for each weekend. The headline prediction for both weekends is that Labor would get 61.5 per cent of the national two-party preferred vote. The Coalition would get 38.5 per cent. On primary votes, the most recent predictions are 51.5 per cent for Labor and 33 per cent for the Coalition.

Bollocks!

There is no way that Labor would get 61.5 per cent of the national two party preferred vote, either last weekend — were an election held then — or in October/November this year when the next election is likely to be held.

Unquestionably Labor is well ahead of the Coalition in the polls. But the margins predicted by Morgan are an overstatement — perhaps a gross overstatement. If we look at both Labor and Coalition primary votes (below), up until the Rudd period we can see a pro-Labor systemic bias in the Morgan series compared with Newspoll and ACNielsen. Since Rudd’s elevation, the one or two point bias has expanded out to three or four points. These biases flow through to Morgan’s two-party preferred predictions.

I have discussed Morgan’s polling accuracy previously. Simon Jackman’s work on the lead-up to the last election that suggested that Morgan under-estimated Coalition support by 4.7 percentage points. In my analysis of the accuracy of polling predictions over the last three federal elections, I found that ACNielsen was within two percentage points 74 per cent of the time. Newspoll was within two points 48 per cent of the time. And Morgan achieved a prediction within two percentage points 22 per cent of the time. If we use a three percentage point margin, the scores improve to 96, 78 and 50 per cent respectively.

Well enough of my cynicism. Let’s return to the optimism of the pollster …

“There are a number of reasons why Kevin Rudd and the ALP should be pleased with the result of this Morgan Poll. Firstly, despite Kevin Rudd admitting to making a mistake in meeting former WA Premier and disgraced lobbyist Brian Burke, ALP primary support is down just 1% to 51.5% the weekend following the revelations.

“Secondly, Prime Minister Howard’s attempt to prolong his attack on Rudd by asking Human Services Minister Ian Campbell to resign for meeting with Burke last year appears to have failed to have an affect on voting intention.

“Thirdly, electors who think the ALP (47.5%) is going to win the Federal Election once again outnumber those who think the L-NP (39.5%) will retain Government. The current ALP “think will win” figure (47.5%) is 19% above the 28.5% that thought the ALP would win in the last Morgan Poll as Kim Beazley as Labor leader.

“On a two-party preferred basis the ALP holds a significant 23% lead: ALP 61.5%, L-NP 38.5%. If an election had been held during the last fortnight, the ALP would have won easily.

“Everyone knows a week is a long time in politics. While the ALP will be pleased with today’s result, much can happen between now and when the election is held at the end of the year.

“As I’ve pointed out on many occasions, the election will be fought on who is best to manage the economy — interest rates, taxation, as well as IR — not who talks to Brian Burke!”

What more can I say? The usual opinion poll graphs are here. You may need to hit the refresh or reload button on your browser to see the latest graphs.

Centrebet has opened a book on the marginal seats in the NSW Election. It gives an insight to how the punters think some key seats will go.

Balmain

Labor is the favourite over the Greens, with the odds giving a winning probability of 56 per cent to 37 per cent

Dubbo

Is neck and neck, with the independent (Fardell on 49 per cent) slightly ahead of the Nationals (48 per cent). Updated 8/3: 57 to 41 per cent in favour of the independent

Manly

Close - the independent (Barr on 50 per cent) is the favourite over the Liberals (44 per cent). Updated 8/3: 50 to 45 per cent

Monaro

Close - Labor (51 per cent) is the favourite over the Nationals (47 per cent). Updated 8/3: 61 to 37 per cent

Murray-Darling

Close - Labor (53 per cent) is the favourite over the Nationals (45 per cent)

Penrith

Labor (64 per cent) is the favourite over the Liberals (35 per cent)

Pittwater

The Liberal candidate (55 per cent) is the favourite over the independent (McTaggart - 34 per cent)

Port Stephens

Labor (57 per cent) is the favourite over the Liberals (40 per cent). Updated 8/3: 59 to 37 per cent

Tamworth

Close - the National candidate (46 per cent) is the favourite over the independent (Draper - 44 per cent).

Terrigal

The Liberal (58 per cent) is the favourite over Labor (38 per cent)

Tweed

Close - Labor (50 per cent) is the favourite over the Nationals (45 per cent)

Wollondilly

Close - Labor (47 per cent) is the favourite over the Liberals (46 per cent). Updated 8/3: 52 to 41 per cent

If you do not want to outlay money, you can enter my (free) tipping competition on the NSW Marginal seats.

Overall, Labor is the clear favourite to win the poll.

The latest Newspoll results were in today’s Australian. If an election were held last weekend, Newspoll predicted that Labor would have won 57 per cent of the national two-party preferred vote. The Coalition: 43 per cent. A landslide win for Labor in anyone’s books.

The attitudinal polling for Rudd appears to have taken a hit following the Government’s intense attack on his integrity and experience. Rudd’s satisfaction rating has fallen by six points and his dissatisfaction rating has gone up by seven points. Nonetheless, the shift to confected vitriol has also cost Howard, with a fall in his satisfaction rating of two points, and an increase in his dissatisfaction rating of three points (over a much lower base than Rudd). Howard is moving into net dissatisfaction territory.

Labor supporters will find vindication in the two-party preferred result, and Coalition supporters will be comforted that nation’s romance with Rudd has peaked and is now on the wane. Who is right? It is simply too early to tell.

In previous election years, the possibility of a Coalition trajectory to victory only became evident in the polls around the middle of the year. Howard is playing the medium term game. He will want to engineer a series of attacks between now and the election — like the Brian Burke attack of the last week — that slowly grind away at Rudd’s appeal with the voters. If successful, come the end of the year, Howard will be in an election winning position.

At this point I should note that according to the Australian,

“The 57 per cent support, compared with the Coalition’s 43 per cent, is the highest recorded in Newspoll surveys since 1993 and is much higher than the two-party-preferred support during the term of the Keating Labor government.”

However, my notes (which I could not cross-check on the Newspoll website) had Beazley’s Labor on 57 per cent in the Newspoll on the weekend of 11 March 2001. That is almost six years ago to the day. Considering the final outcome, it is not a good omen for Rudd.

The usual opinion poll graphs are here. You may need to hit the refresh or reload button on your browser to see the latest graphs.

One final observation: Howard has accused Rudd of inexperience and not understanding our system of government with his call for an election now. Constitutionally, a House of Representatives election prior to August 2007 would necessitate a separate half senate election somewhere between August 2007 and May 2008. Ironically, Menzies — Australia’s longest serving Prime Minister — called an election in 1963 at the height of his experience that saw a series of subsequent separate half senate elections over the following decade.

Elsewhere: The Poll Bludger gets stuck into the spin coming out of the NSW ALP machine.