Can we trust the polls?

Fairfax-Neilson polling has shown that Prime Minister Kevin Rudd has failed to improve the Labor Party’s standing in the crucial state of Queensland prior to this Saturday’s Federal Election. Seven out of 10 Queensland residents surveyed think that Tony Abbott will be the next Prime Minister. Labor is trailing the Coalition 47 per cent to 53 per cent on a two-party preferred basis, while Clive Palmer’s Palmer United Party (PUP) is polling as strongly as the Greens, with the primary vote at eight per cent (Labor is only at 31 per cent).

But recently media outlets have been questioning both the relevance and the accuracy of the polls. It has been noted that, although 70 per cent of Queenslanders expect Labor to lose the election on Saturday, it isn’t because they object to Rudd as Prime Minister. Almost two-thirds of Queenslanders polled stated that they had already decided how they were going to vote before Julia Gillard stepped down as Prime Minister in June, even though the numbers have been attributed to Rudd.

But the biggest flaw in the polling system is that millions of voters are being ignored. One-in-five Australians does not use a landline – they are mobile phone-only users. The number of people who only use mobile phones has grown by 20 per cent in the past year alone, and both Neilson and Newspoll are ignoring these voters.

Most mobile phone-only users are city-based renters under 35 years of age. Other mobile phone-only users include retirees travelling around Australia. EssentialResearch, which allows these Australians to be included in the polls by running its political polls online, has labor polling at 50 per cent, a significantly better result for Labor than either the Neilson or Newspoll have gathered. Andrew Bunn, research director at EssentialResearch has explained that, “On average, the most accurate polls in the United States for the last Presidential Election were online”. On the same US election, statistician Nate Silver reported in the New York Times that telephone polls which only called landlines “performed poorly and showed a more Republican-leaning electorate”.

Find out more about the recent Fairfax-Neilson poll on the Canberra Times website. 

Read more about phone polling on the Sydney Morning Herald website. 

Living in the dark ages

I am one of the lucky ones. Unlike most Australians, I have not suffered through the pre-election media hype. I managed to time my annual leave to coincide with the lead up to the election, so I have only been dealing with the dog fight which is ‘Federal Election 2013’ for a few days.

On the downside, I did feel pretty out of the loop when I returned to Australia. So I turned to the polls to find out who Australia will be voting for in four days’ time. We’re not supposed to say it (not yet, anyway), but if the polls are as accurate as we’ve been led to believe then the Coalition will win the election on Saturday. I say the Coalition, not Tony Abbott, because even though Rudd and Abbott have turned this election into a popularity contest we do actually vote for a party, not a leader, in Australia.

But it turns out that the two biggest pollsters in Australia, Neilson and Newspoll, are only calling landlines. Why? Because door-to-door polling costs more money. And putting together accurate mobile-phone calling lists costs money. Which makes me feel pretty cheated.

I rely on the Bureau of Meteorology (BOM) to give me an accurate weather forecast. If the BOM started to give Australians bad weather information because it was cutting costs, I would expect most people to be upset by that. In the same way I, like many Australians, am relying on the polls to give an accurate forecast of this year’s Federal Election. And I am upset that this year the polls might be giving me bad information because Neilson and Newspoll refuse to understand that the landline is no longer the only form of communication.

Those being left out are the under 35s. Why should this matter to YOURLifeChoices members, other than as a matter of principle? Because it affects both parties. Statistics show that the under 35-year-olds tend to be left leaning, which in Australia means giving their votes to Labor instead of the Coalition (I’m working on the safe assumption that the Greens are not set for a landslide victory this year). By leaving these voters out of the polling, poll results show an inaccurate right-wing bias, as seen in the recent US election poll results. This hurts both parties. Labor supporters become dispirited, or swing voters feel they should follow the masses according to the polls. The Coalition supporters think that they’re safe, and don’t campaign as hard as they might otherwise.

Why then are newspapers still relying on expensive, inaccurate poll results, and dragging us down with them? I think it’s time that Australia embraced the digital age and online polling. Real Australians are tech-savvy, and it’s time that Australian politics caught up.

Do you trust the polls? And do they affect how you vote? Or have you already made up your mind?