Important messages from the polls

The AFR Fairfax-Ipsos page helpfully reminds us that Kim Beazley gained 50.98% of the vote in the 1998 election but did not win. So with the two-party preferred vote now at 51-49 to the LNP, officially it’s too close to call. Anyway that slight lead is offset by Newspoll which came in at 51-49 to Labor for the fourth time in a row. So a couple of weeks of electioneering appear to have made no difference overall. Yet there are, I think, some important messages to be mined from the polls.

Firstly, the voters in Queensland are grumpy. Fairfax-Ipsos does a state by state breakdown showing the LNP and Labor level in NSW and SA, in Victoria and WA Labor is a bit in front. In Queensland, however, the LNP leads by a whopping 58-42. At that rate probably no seats would change hands, making it difficult for Labor to win nationally.

Labor currently holds 55 seats in the HoR, and needs around 20 to form government. Currently Labor holds 6 out of 30 in Queensland. The marginal seats list shows it needs six more in Queensland. If Fairfax-Ipsos is right it will win none.

The poll has a margin or error of 5.7% for the Qld result, but you’d expect the results to cluster near the middle. Poll Bludger reported on 14 May that federal voting intentions in Qld favoured the LNP 54-46, as against 57-43 at the last election. On that basis only Capricornia and Petrie would go.

I think the basic problem is the economy in Qld which is cactus.

Secondly, the Greens vote, 8.7% at the last election, has consistently been at 14% for Fairfax-Ipsos and 11% in Newspoll. Essential has them at 10% and Roy Morgan at 15.5%. At the very least these polls augur well for their senate voting.

Third, PUP is gone. Newspoll doesn’t ask, but Fairfax-Ipsos has PUP at 1%.

Fourth, there is a trend for Turnbull’s personal standing to sink while Shorten’s improves. On the role performance question, Newspoll has them both on a net -12. Fairfax-Ipsos has Turnbull at +10, but sinking from +23 at the beginning of March. Shorten is at -6 compared to -29 at the beginning of March.

In Fairfax-Ipsos on preferred Prime Minister Turnbull has gone from a lead of 39 over Shorten to 17 over the same period. In Newspoll Turnbull’s lead has shrunk from 34 to 15.

If anything personality is receding as a factor.

Fifth, and last, Fairfax-Ipsos asks voters who they think will win. This is said to be a better predictor than the voting intentions poll. Here the punters backing a LNP win have jumped to 57% compared to 53% in the last poll.

Laura Tingle thinks this could be deadly for Turnbull. If voters are so confident he will win they might feel safe to give him a kick in the shins, leading to an ‘accidental’ election result.

Tingle also says people are not paying attention, seeing the electioneering as nothing more than “annoying background noise”.

Finally, Tingle reports an aspect of the poll that didn’t appear in the paper. Voters would vastly prefer money for schools over tax cuts for business:

Education v tax cuts_cropped_600

They should know who to vote for, then, shouldn’t they?

BTW, Kevin Bonham thinks Labor needs 50.9% of the vote for a 50:50 chance of winning.

4 thoughts on “Important messages from the polls”

  1. Douglas, thanks for that. Before I posted I checked whether Adrian Beaumont had done a piece. If the time stamp is to be believed I missed it.

    He asks the question Is the “Think will win” question relevant? He says:

    I do not agree with those who assert that undecided voters will vote for the party they think will win.

    That’s what he says is more relevant in voluntary voting countries. I didn’t say that, Laura Tingle didn’t say that, no-one I know said that!

    Beaumont eventually says:

    If people think one party will easily win, but the polls are close, there is a danger that the other major party will benefit from protest voters, and surprisingly win.

    Which is exactly what Laura Tingle was saying.

    He goes on to say:

    That danger also applies to betting market expectations. Betting markets have not performed well at many state elections, and I disagree with those who talk about the wisdom of the betting markets.

    Which is contra what Tingle said. It actually agrees with my own hunch, FWIW.

    Adrian Beaumont has an advantage over me in that he obviously has access to the full poll. I didn’t know that Newspoll had asked that question, although I had driven 2km to buy Uncle Rupert’s rag which I don’t normally subscribe to.

  2. Impressed by your commitment to seeking information – even to the extent of purchasing a Murdoch rag. The difference between the two polls on that question is interesting.

  3. Douglas, I used to subscribe to the Australian because it’s quite good on financial affairs and sport, but I objected to paying for an entity that was an active player in the political game. Usually I can pick up one at any of a couple of foodstores I’m likely to drive past, but they’d sold out, so I had to head down to the newsagent.

Comments are closed.