Inside Story

Who’ll win, and by how much

Brian Costar does the maths – and it’s not a repeat of 1931

Brian Costar 19 August 2010 1153 words

The Australian Broadcasting Company’s 1931 federal election broadcast team in action in an emergency studio - the kitchen of Harry Croot’s home in Arncliffe. Norman Lyons, Talks Officer, is taking figures over the phone, while Harry Croot helps with the tally sheet. Australian Broadcasting Corporation

LESS THAN ten months ago the Rudd Labor government was cruising towards an electoral victory with a moderate increase in its House of Representatives majority. Bold policy initiatives had seen off the global financial crisis, the Coalition was a rabble and Labor was aided by favourable electoral redistributions in the key states of New South Wales and Queensland.

Yet by early 2010 the government had “lost its way” and, in an unprecedented move, the prime minister was replaced by his deputy. Those of us who do not inhabit the Canberra beltway soon discovered that the decision was a wise one. Put bluntly, for all his obvious qualities and talents, Kevin Rudd didn’t have what it takes to be a good prime minister and, in Winston Churchill’s memorable words, needed to be “pole-axed.”

Some have doubted the wisdom of Julia Gillard’s decision to call an election so soon after her elevation, but the Australian economy remains the envy of the developed world, with debt, unemployment and inflation all at comparatively low levels. Why then is the re-election of the government still doubted? After all, few people who are still alive had a chance to vote last time a first-term government lost an election, way back in 1931.

Partly, it’s because of the growing disconnection between the “political class” and the electorate as a whole. Press gallery journalists in particular react to every morsel of policy announcement, every perceived gaffe by the politicians and every microscopic movement in opinion polls as earth-shattering events; ordinary people know better.

That this campaign has been boring has become a cliché. But we’d better get used to it, because the parties and their campaign managers are so locked into the 24/7 media cycle that never again will we see leaders putting forward bold and integrated policy manifestos. Instead they will drip feed the media with disconnected policy statements and thirty-second sound bites and, above all else, they will stay on message (even though they don’t have one) and be risk-averse.

I suspect that people see through all of this and vote according to their perceptions of the parties’ likely performances in the key areas of economic management, the environment, health, education and social well-being – the big-ticket items of federal electoral politics.

So what is likely to be the result of Saturday’s election? Perhaps because of boredom with the campaign, the spectre of a hung parliament has been raised. Predicting an outcome in which neither major party has a majority on the floor of the House of Representatives is like predicting a draw in the AFL Grand Final – it’s possible but highly unlikely. The last federal election that delivered us a hung parliament occurred in 1940.

All the party operatives and most journalists tell us that the result will be “very close” and will “go down to the wire” – as they always do. Australian elections are always close on the votes and rarely so on the seats. The last really close election was in 1961, when the Coalition secured a bare majority of two seats.

Most commentators are helpfully telling us that the election will be won in the marginal seats, a proposition that Peter Brent of Mumble reminds us is nothing more than “the bleeding obvious.” It is worth remembering that marginal seats are the product of socio-demographic residency patterns corralled into electorates by boundary drawing. The inhabitants of those electorates are not essentially different in their beliefs from other Australians; they are not strange creatures recently arrived from the planet Adenoid.

In the last week we have been offered a couple of marginal-seat polls, which, despite some overblown promotion, need to be treated very cautiously. Political-polling newcomer Telereach surveyed over 28,000 voters in fifty-four marginals, which looks impressive until you realise that the sample per electorate – in some cases as few as 400 – produces a margin of error of around 5 to 6 per cent. Since marginals are defined as those held by 5 per cent or less, this doesn’t help us very much.

Even less useful was a Galaxy poll that had a sample of only 200 in each of twenty marginals and a margin of error off the scale. It’s possible to do reliable marginal seat polling but no one will commit the funds needed to do so – except the big parties, which won’t tell us their results.

Regional variations are said to be a determining feature of Saturday’s result. The pollsters seem to agree that Labor is well ahead on two-party-preferred votes in Victoria and South Australia, just ahead in New South Wales, just behind in Queensland and well behind in Western Australia. The samples are too small to be definitive about Tasmania but Labor will probably hold all of its five seats.

Western Australia has a reputation for defying national voting trends. In the Labor rout election of 1966 there was a swing of 3 per cent to the party in the west. This time Labor will be fortunate not to lose the seats of Swan and Hasluck. South Australia, meanwhile, is Labor’s strongest state, with the seats of Sturt and Boothby likely additions.

For years now Victoria has been the jewel in Labor’s crown. The party should easily pick up McEwen and Latrobe and be an outside chance in Dunkley. But it is almost certain to lose Melbourne to the Greens.

New South Wales is a key state and historically has been strong for Labor. Federal Labor is defending a clutch of very marginal seats and is not being helped by the dead government walking in Macquarie Street. The Coalition should win Macarthur, Macquarie, Robertson and Gilmore and have some chance in John Howard’s former seat of Bennelong.

My old home state of Queensland is this election tipster’s nightmare. Favourite son, Kevin Rudd, wowed them in 2007 and Labor scored some spectacular double-digit swings to win an additional nine seats. Queensland, however, has a history of being very pro-Labor at one election and rejecting it at the next.

Both parties are defending a significant number of very marginal seats: eight for Labor and six for the Liberal National Party. To further complicate matters, Queensland is still a quite decentralised state with a lower proportion of its population living in the capital city than any of the other mainland states. It is just possible that Labor will lose the regional/quasi-regional seats of Herbert, Flynn, Dawson, Forde and Leichhardt, but retain Dickson and Longman and win Bowman and Ryan in Brisbane – or any combination of the last four.

What then will be the result? A Labor victory, certainly, but by what seat margin?

Best case for Labor
Labor: 81
Coalition: 65
Others: 4

Worst case for Labor
Labor: 78
Coalition: 68
Others: 4