It's the Marginals Stupid

After a year in which the opinion polls have pointed unfalteringly to a Coalition wipeout, the Prime Minister and Opposition Leader have adopted tellingly different strategies to manage expectations.

Vigilant against charges of overconfidence, Kevin Rudd’s mantra not too strong a word, given the frequency with which he utters it is that Labor faces a metaphorical Mount Everest to win the 16 seats needed for a majority. It is possible to find evidence supporting this case, providing it is chosen selectively.

While Labor would outpoll the Coalition in two-Party terms with an overall swing of 2.8 per cent, a uniform swing of that size would deliver only 12 seats two of them in Western Australia, which is said to be bucking the trend so heavily that Labor might go backwards there. It can further be argued that Labor will achieve redundant swings in safe Labor and Liberal seats, where voters will respectively be swayed by WorkChoices and post-materialist ‘doctors’ wives’ issues, but fall short where it matters most.

Thanks to Fiona Katauskas

Senior figures in the Coalition have been arriving at a similar story from a different angle. Their challenge has been to make re-election seem a credible prospect, thereby encouraging the troops to stay keen and persuading business donors to open their pockets. ‘Liberal sources’ have thus been heard talking of Party polling showing crucial marginals not following the overall trend.

One highly publicised example was John Howard’s attempt to quell Party room nerves during the early September leadership crisis. The Prime Minister spoke of encouraging intelligence from Eden-Monaro, the renowned ‘bellwether’ electorate which has gone with the winning Party at each election since 1972. However, it’s hard not to be suspicious of such talk when the Liberal hierarchy reportedly keeps the hard facts of its polling confidential at the highest level.

Similarly, Kevin Rudd’s ‘Everest’ metaphor only holds true so long as one views the 2004 election as a fixed starting point. That election pushed many traditionally marginal seats high up the Coalition side of the MacKerras pendulum. In Queensland alone, Petrie, Herbert, Hinkler, Longman and Bowman were all considered key seats going into the 2004 election, but now sit beyond the marginal seat range as defined by the Australian Electoral Commission (which states: ‘ By convention, when a Party receives less than 56 per cent of the [two-Party preferred] vote the seat is classified as ‘marginal;’ between 56-60 per cent it is classified as ‘fairly safe’ and more than 60 per cent is considered ‘safe’).

In Victoria, in 2004, the Liberals recorded easy wins in the traditionally knife-edge seats of Deakin, La Trobe, McMillan and McEwen. There were particularly large swings in the mortgage belt, where voters renowned for weak Party loyalties and sensitive hip-pocket nerves responded strongly to the Coalition’s campaign on interest rates.

The published polling gives ample reason to think these seats are back on the table, with Labor on track for near double-digit swings in every State except WA.

Sure enough, there are a number of marginal seats where the Liberals have reportedly given up so that sparse campaign resources can be allocated elsewhere. These include Kingston, Hindmarsh and Makin in Adelaide, each held by margins of less than 1 per cent.

The Liberals are also up against it in the inner Brisbane seats of Bonner and Moreton, the former of which was a surprise win in 2004.

In Sydney, Labor should have no trouble picking up the 0.8 per cent swing needed to retain Parramatta, made notionally Liberal by the redistribution. They should at last recover the Penrith-based seat of Lindsay, buoyed by the retirement of Jackie Kelly and a redistribution that has cut the margin from 5.3 to 2.9 per cent.

Elsewhere in New South Wales, Labor are favourites to win the Central Coast seat of Dobell and, despite the Prime Minister’s protestations, Eden-Monaro. A number of coastal seats north of Sydney with solid Coalition margins are also in play, including Robertson, Paterson, Cowper and especially Page, where sitting Nationals member Ian Causley is retiring.

Sydney offers the twin spectacles of Bennelong and Wentworth, respectively held by the Prime Minister and one of his aspiring successors. While it is tempting to imagine that the mantle of leadership will protect John Howard in Bennelong, there is no objective reason to think it will swing differently from any other middle-Sydney seat. Howard is likely to go down with the ship if his Government is defeated, which at the moment looks a much better than even-money prospect.

Malcolm Turnbull’s margin in Wentworth is even narrower, having been cut from 5.6 per cent to 2.6 per cent by the addition of inner-city Paddington and Darlinghurst in the redistribution. However, Wentworth is more affluent than Bennelong and less sensitive to mortgage pressures. The newly added areas are slightly less weak for the Liberals than the 2004 election results suggest, given that they ‘played dead’ that time in the hope of helping the Greens beat Labor in the neighbouring seat of Sydney.

These areas will now experience the full force of the well-oiled Turnbull campaign machine.

The Liberals are also campaigning like they mean it in the blue-ribbon seats of North Sydney (held by Workplace Relations Minister Joe Hockey) and Ryan in Brisbane, though both are likely to hold.

Victoria is notable for its lack of low-hanging fruit, the smallest Liberal margin being 5.0 per cent in Deakin. Nonetheless, any seat with a margin of less than 10 per cent cannot be regarded as bolted down, including Peter Costello’s seat of Higgins (8.8 per cent).

A more realistic Labor hit list would include Deakin, La Trobe, McMillan, McEwen and especially Corangamite, outside Geelong.

To Labor’s three certain Adelaide gains can be added the possibilities of Sturt and Boothby, although the poor performance of pseudo-celebrity candidate Nicole Cornes appears to be damaging them in the latter.

The Coalition reportedly hopes to limit the damage in Queensland to Bonner and Moreton, although most would consider that optimistic. No expense is being spared to shore up support in the Ipswich-based seat of Blair, explaining the Government’s contentious decision to fund the Goodna Bypass as its preferred solution to congestion on the Ipswich Motorway.

Labor is confident about a number of seats further north, including Townsville-based Herbert, the new seat of Flynn and further north Leichhardt, where the retirement of local legend Warren Entsch will bite deeply into the 10.3 per cent margin from 2004.

That leaves two States and one Territory which are laws unto themselves. Labor is likely to win the Darwin-based seat of Solomon, although the Party’s well-placed former pollster Rod Cameron has included it in a list of marginals that can’t be taken for granted.

It was long thought that Labor would win a clean sweep in Tasmania by recovering Bass and Braddon, which were lost in 2004, but published polling suggests Government largesse has at least made the Liberals competitive in Braddon.

With Western Australia said to be bucking the national trend on the back of its resources boom, the retirement of popular Labor member Graham Edwards has the Liberals eyeing his seat of Cowan as possible compensation for a loss elsewhere. This has been backed by a recent poll in The West Australian showing the Liberals ahead in Cowan and their own Perth marginals, Stirling and Hasluck. However, this appears at odds with other polling pointing to a Statewide Labor swing in WA of between 3 and 5 per cent.

Despite the polls’ remarkable consistency throughout the year, many observers have refused to take them at face value. It was argued that the announcement of the election would produce a ‘narrowing’ as voters finally focused their minds on the Government’s strong economic record.

Early campaign polls have provided only limited support for this thesis. Recent ACNielsen, Roy Morgan and Galaxy surveys have removed some of the gloss from Labor’s lead, while still showing them on track for a commanding win.

Any suggestions that the Coalition was building momentum were silenced by Labor’s stunning 58-42 lead in last week’s Newspoll although this week’s 54-46 result suggests this may have been an aberration. Crucially, much of the variation in the TPP vote has been an artifact of fluctuating minor Party preferences. Labor’s primary vote, it seems, is set firmly in the election-winning high 40s.

In summary, local variations are always hard to predict, but a rough interpretation of the State-level picture points to around six Labor gains in each of New South Wales and Queensland, and four each in Victoria and South Australia. Labor is also very likely to win Solomon, Bass, Braddon and either or both of Stirling and Hasluck.

With independents headed for certain re-election in Kennedy and New England, that points to an overall outcome of around 84 seats for Labor in a chamber of 150, against 64 for the Coalition.