Canning a psephologists’ nightmare

Don't be fooled by those claiming to be able to predict the winner of the by-election for the federal seat of Canning. Even after four weeks of frenetic campaigning by the major parties, the truth is that nobody really knows.

The conventional wisdom is that, with the late Don Randall winning for the Liberal Party in 2013 by a healthy majority of 11.8 per cent, former SAS captain Andrew Hastie will retain the seat for the government.

That’s a reasonable assumption, but several variables need to be considered.

The first is that Mr Randall had built up a significant personal vote since winning Canning for the first time in 2001. Labor threw everything at him in the 2010 poll, when former state minister Alannah MacTiernan was his opponent. But even she couldn’t budge the maverick Liberal, who won with a margin of 2.2 per cent.

At the past two federal polls, support for federal Labor in Western Australia was at its weakest point for years. The administrations led by Kevin Rudd and Julia Gillard succeeded in alienating big sections of WA voters, with the mining tax in particular a spectacular vote loser in resources-rich WA.

Relations between state Labor MPs and the federal Labor government also slumped, and corporate campaign donations to the party virtually dried up. The Liberals were the delighted beneficiaries. To say federal Labor was out of step with the aspirations of most Western Australians would be an understatement.

So the Liberals had a well-resourced campaign, and Labor was struggling for funds ahead of the 2013 poll. Labor did not rate Canning as a high priority seat, let alone winnable, in 2013. So its vote slumped further.

The story is different this time.

Another factor helping Mr Randall in 2013 was the ballot paper order. While all candidates prefer to be top of the list to benefit from any donkey votes (where voters place ‘1’ against the top candidate and fill in preferences down the page), the Liberal MP’s margin was so healthy the ballot paper position was irrelevant. Yet Mr Randall drew top spot, which no doubt boosted his vote further.

The two serious contenders this time had no such luck. Mr Hastie drew the sixth position, and Labor’s Matt Keogh, is in eighth spot. Any donkey vote will favour Mr Hastie.

The rapid population growth in Canning, which includes the two state Labor seats of Armadale and Mandurah, but also extends south to the rural areas around Waroona and east to the mining sites in Pinjarra and Boddington, provides another variable to the uncertainty.

Canning takes in the rapidly growing south-east metropolitan corridor. At the height of the resources boom, the migration component of WA’s population growth was assessed at about 50,000 annually. That’s 1,000 a week; and many were accommodated in the new sub-divisions in Canning.

The seat had 101,804 registered voters in 2013. When the roll closed for the by-election, this number had swollen to 112,809. Who are these additional voters? Are they mainly home buyers, renters or on social security?

The answers could be all important to the result.

And how many voters on the roll will actually go to the polling booths. Voting is compulsory, of course, but only 92 per cent of those on the roll turned up last time.

It’s a challenge to get the disaffected to participate in a by-election, despite the best efforts of the Australian Electoral Commission and the parties. Even though the result could have a significant impact on politics over the next 12 months, especially for the personal fortunes of Prime Minister Tony Abbott and opposition leader Bill Shorten, it’s possible less than 85 per cent of those eligible might actually vote.

Would the absentees have supported Labor or Liberal? No-one will ever know.

Extraordinary claims have been made during the campaign. One was that Labor was ‘running dead’, and wanted the Liberals to win so that Mr Abbott would remain prime minister, making a Labor win more likely in next year’s general election. Fanciful.

Another was that Mr Hastie was distancing himself from Mr Abbott, despite them campaigning together next day. Fanciful again.

So predicting the result is extremely imprecise. However it’s reasonable to assume that – short of an unexpected collapse in Liberal support – closing a margin of 11.8 per cent will prove too much for Mr Keogh.

The result will be decided on preferences, nothing surer; and that means it might not be known for several days. The Liberals will say a win’s a win, while Labor will claim a swing of more than 7 per cent to be as good as a win.

But should Mr Keogh gain the desired swing, Labor will be elated. And Mr Abbott’s future as prime minister will be very tenuous indeed.