Predicting an election result is a mug’s game. The major parties run tracking polls that give them a reasonably good idea of how they are faring on a seat-by-seat basis. But they keep them under wraps. They background journalists – making claims that might be accurate but might also be spin.
For three weeks Coalition sources have been briefing that it’s all over bar the shouting, that they will lose fewer than 10 seats. That could be a true reflection of their polling, or it could be a strategy to deny Labor any final week momentum and help Malcolm Turnbull’s message contrasting the “stability” of a clear Coalition majority with the potential “chaos” of voting Greens or independent.
But the Australian electorate has also shown signs of volatility and general disaffection with both major parties. More than 20% of the Australian electorate intends to vote for the Greens or a minor party and another 7.2% remain undecided. Turnbull has been warning against a vote for Greens or independents precisely because he’s worried so many disillusioned voters are intending to. With many electorates very close, strategists on both sides concede things could still shift before polling day.
If we adjust for the redistributions since 2013 (in which one Labor-held seat was abolished in New South Wales and a notionally Liberal seat was created in Western Australia, while three other Liberal seats in New South Wales have become notionally Labor) and allocate Clive Palmer’s seat back to the Coalition, Turnbull would lose his majority if he loses 14 seats and Labor would gain a majority if it can win 19.
Guardian Australia’s poll aggregation of national polls puts the Coalition on 50.35% of the two-party preferred vote. William Bowe, on his Poll Bludger blog, puts the Coalition on 50.2%. This represents a swing against the Coalition since the last election of about 3.3%. If that was uniform across the nation, the Coalition would lose 13 seats.
But, of course, the swing is never uniform, so that result could also be refined by aggregating the polls state by state.
Bowe helpfully does this and each state’s swing can be compared with the marginal seats in that state.
In New South Wales, the state swing is 3.6%, which, if uniform across the state, would see the Coalition lose six seats – Banks, Eden-Monaro, Lindsay, Page, Robertson and Reid – on top of Dobell, Barton and Paterson, the seats that have become notionally Labor after the redistribution.
In Victoria, the state swing is 3%, which, if uniform, would cost the Coalition just one seat: Deakin.
In Queensland, the swing is 4.8%, which would take out five seats: Petrie, Capricornia, Bonner, Brisbane and Forde.
In Western Australia, the swing is 7.5%, which would cost four seats: Cowan, Hasluck, Burt and Swan.
In South Australia, Tasmania and the territories, it suggests no seat changing hands.
That adds up to a Coalition loss of 16 seats, which could mean a hung parliament. Bowe applied a separate probability calculation to each seat and comes up with a loss of 12.
But swings also vary between seats in the same state, according to demographics, local issues, the performance of the sitting member, the strength of the opposing candidates and, perhaps, the big spending promises for local sporting grounds and CCTV cameras that both parties have made.
That means looking at the published seat-by-seat polling (often less accurate than national polling but at least some indication), the itineraries of the leaders (usually a very good indication of which seats are in play) and the specific state issues, as well as what the strategists say.
In NSW, the Coalition insists its only possible loss is Eden-Monaro, beyond the three seats that became notionally Labor under the redistribution: Barton, Paterson and Dobell.
But Turnbull was campaigning just last Monday with the sitting member for Lindsay, held by the Liberals by 3%. Four polls earlier in the campaign showed the Liberals leading in that electorate but the most recent poll by the NSW Teachers Federation put Labor well ahead. Shorten held his campaign launch in the seat and has also made multiple visits.
Looking at what the parties are doing, as well as what they are saying, would suggest Lindsay is still in play, as well as the northern regional seat of Page, where two recent polls put Labor’s candidate in the lead but the Nationals are pouring in resources. The teacher’s union poll also suggested Labor was in with a chance in Macquarie and Gilmore. The Liberal launch this Sunday will be in Reid, which doesn’t mean the Coalition is worried about that particular seat but does signify the importance of western Sydney.
In Victoria, the strategists say it is unlikely Labor will lose any seats. But it is unclear how the bitter firefighters dispute will influence the outcome in several marginals. And Turnbull was still campaigning in Corangamite earlier this week, a seat where one Newspoll and two Reachtel polls have showed Labor in the lead.
We’re also told no seats are likely to shift in Tasmania but Turnbull was there on Friday, matching a Labor education spending commitment, so he can’t be entirely confident, and other sources say Lyons and Bass are both very close.
Shorten made another dash to Western Australia this week, underlining Labor’s high hopes in Burt, Cowan and Hasluck. And the Labor leader was in the Northern Territory on Friday because Labor is still hoping to win Solomon.
Both sides acknowledge Labor isn’t getting the swing in Queensland it had been seeking, but Labor remains hopeful about Capricornia, and thinks it is a chance in Dawson and an outside chance in the Brisbane seat of Forde. Labor’s hopes for the hyper-marginal seat of Petrie are fading.
Which leaves South Australia, where the surging Nick Xenophon Team is causing both major parties headaches. Polls indicate the NXT could win Mayo and possibly even Grey and Barker.
But Labor is worried about the Melbourne seat of Batman, where another poll this week suggested frontbencher David Feeney could lose to the Greens. Polls have given hope to the independents Tony Windsor and Rob Oakeshott, trying to re-enter parliament in New England and Cowper, but independent Bob Katter could lose his seat of Kennedy to the Nationals.
That adds up to too many variables to make a clear prediction, even before we add the fallout from the Brexit economic uncertainty to the equation. A Labor victory looks very difficult, a narrow Coalition victory looks more likely, and a hung parliament is at least possible.
.jpg?w=600)
