The Australian has published its quarterly geographic and demographic breakdowns of its federal polling data, compiling the results of the four polls it has published this year (it took January off). The national figures are therefore no surprise to us, as they are merely an average of this year’s polling: Labor on 34 per cent of the primary vote compared with 38.0 per cent at the election, the Coalition on 42 per cent compared with 43.6 per cent and the Greens on 13 per cent compared with 11.8 per cent (remembering that phone pollsters seemed to have acquired a tendency to overrate the Greens and underrate Labor). The Coalition holds a two-party lead of 51-49, compared with 50.1-49.9 to Labor at the election.
The real interest in the figures is in the various breakdowns offered, particularly by state. The most distinctive result on voting intention is the solid recovery for Labor in Western Australia, from a dismal base of 43.6 per cent at the federal election to 45 per cent in October-December 2010 to 48 per cent this time. A Labor hike in Queensland from 44.9 per cent at the election to 48 per cent late last year looked rather too much at the time, and sure enough the latest poll has it moderating to 46 per cent. Labor’s decline overall has been driven by NSW/ACT, from 49.5 per cent two-party at the election to 48 per cent in both quarters, and Victoria, from 55.3 per cent at the election to 55 per cent to 53 per cent. In South Australia, where Labor has nothing in the way of tight marginals, they have gone from 53.2 per cent at the election to 51 per cent and 52 per cent.
If such swings were uniform, the Coalition would gain Greenway, Robertson, Lindsay and possibly Banks in New South Wales, plus Corangamite and La Trobe (but not quite Deakin) in Victoria. Labor would gain Hasluck, Canning and Swan in Western Australia, and Brisbane in Queensland. Other things being equal, and leaving Banks with Labor, there would be a net shift of one seat in the Coalition’s direction: from 73 Coalition and 72 Labor to 74 and 71. This of course makes the notably unsafe assumption that all sitting cross-benchers would be re-elected. Furthermore, the capitals and non-capitals breakdowns suggest it would be worse for Labor than that. In the metropolitan areas which are home to most of the marginal seats, the two-party vote is at 50-50 compared with 52.5-47.5 in Labor’s favour at the election. In the non-capitals Labor has gained ground, now trailing 52-48 rather than 53.4-46.6.
On personal ratings, the most interesting finding is that both leaders have soured among the 50-plus age group. The results for Tony Abbott defy some of the stereotypes about his support base: his 52 per cent disapproval among the 50-plus is the highest of any age group, and a once substantially higher approval rating among this cohort has fallen right back to the field. He has also lost ground among 35-49s, as has the Coalition on the primary vote. Gillard is down four points on approval and up five on disapproval among the 50-plus, a situation which is reversed among the 18-34s, now clearly her best cohort.
New South Wales and Queensland are about equal as Julia Gillard’s worst state, owing to a post-election recovery in Queensland. Victoria and South Australia are roughly equal as her best (although her disapproval is up in South Australia), with Western Australia surprisingly close behind. Tony Abbott’s ratings have been consistently mediocre in New South Wales and Victoria and consistently neutral in Queensland, but he has weakened considerably in South Australia and Western Australia: from net neutral to minus 14 and minus 12. Gillard’s lead as preferred prime minister is only six points in Queensland, elsewhere ranging from 15 points in New South Wales to 23 points in Victoria. Gender splits lean in the expected directions, though not as heavily as you would think. An exception is disapproval of Gillard, with women notably more reluctant to give her the thumbs down.