Blog CemeteryOpinion Polls Pt 2
November 19th 2007 05:42
QUEENSLAND
Due to the small number of ALP seats in Queensland, and the high Coalition TPP vote in 2004 in the state, the graphs for Queensland assume the safety of three ALP and six coalition seats. This leaves 3 ALP and 16 Coalition seats being analysed.
ALL POLLS
Although results early in the year placed seats such as Fisher and Fairfax at risk, since winter most polls have suggested that five seats (Petrie, Hinkler, Flynn, Bowman and Dickson) would be the most hard-fought in the election. The ALP have always been ahead in Bonner and Moreton, while generally ahead in Blair, Herbert and Longman. While a handful of polls have placed safer Coalition seats such as Forde and Leichhardt at risk the majority would see these seats retained by the Coalition with reduced majorities.
STATE ONLY POLLS
Focusing purely on state polls provides something for everyone. Coalition supporters can highlight polls in early October and mid November which minimized their losses to within two and five seats. Both polls were conducted by Galaxy polling and limited to the marginal seats of Bonner, Herbert, Longman and Moreton. The weakness here for the Coalition is that the small swings in marginal seats may not be enough as the polling ignores other seats which would also fall to the ALP if the statewide polling is correct. Galaxy may hold promise for the Coalition in these marginals but the statewide swing would probably give Labor other seats, such as Blair, Petrie and Bowman.
Further there are also polls which look good for Labor. Every non-Galaxy poll gives the ALP at least six seats, with the very safe seats of Fisher and even Fairfax under threat if statewide swings are evenly distributed across the state. While Galaxy's most recent poll shows the Coalition losing just two seats it was conducted at the same time as a Morgan poll which suggests the ALP would gain all sixteen seats considered here.
VERDICT: Queensland is going to make at least one of Galaxy or Morgan look pretty silly. On the whole the data suggests good gains for the ALP, although the Morgan prediction is a little over-the-top. The Coalition may be able to safe Herbert, Longman, and Hinkler due to local factors but the polls suggest the ALP will gain between six and ten seats in Queensland.
SOUTH AUSTRALIA
ALL POLLS
The graph of the year's national and state polls shows South Australia to be very stable in terms of seats won. Every poll has the ALP favoured to hold their marginals of Adelaide and Hindmarsh while also winning Kingston, Wakefield and Makin from the Liberals. Beyond this, most polls have the ALP winning Boothby and Sturt, although an increasing number have Sturt as too-close-to-call. While a few early polls gave the ALP a chance in Mayo and Grey, only one poll in the campaign moved these out of the Coalition's total.
STATE ONLY POLLS
The statewide figures are one of steady improvement for the Coalition, although at no stage have they come close to holding one of their three marginals and only twice have they been within 1% at Boothby. While in March the state was the Coalition's weakest, with the cumulative Newspoll giving the ALP all but Barker, recent surveys have seen Mayo and Grey return to the Coalition and their position in Sturt improve slowly. The recent Galaxy poll was the first statewide poll, and only the fourth overall, to give the Coalition over 51% in Sturt.
VERDICT: The polls have SA as a very strong state for Labor. Kingston, Wakefield and Makin would require the mother of all local campaigns to not fall to Labor, while Boothby sees Labor favoured and Sturt has them no worse than par with the Liberals. Both seats may be decided by the merits of their respective candidates (in Boothby the Labor candidate has had her share of troubles, while in Sturt the Labor candidate has been touted as a future star but is up against long-term member and Government Minister Christopher Pyne) but it wouldn't be a surprise to see Labor win both. But the only possible analysis of the polls is a three to five seat gain to Labor, with Boothby leaning Labor and Sturt leaning Liberal.
WESTERN AUSTRALIA
ALL POLLS
Due to the large difference between the federal swing to Rudd and the swing recorded in state polls, WA is the one state where looking at the federal graphs is not so helpful. For what it's worth, early polls indicated that Coalition strongholds such as Pearce and Curtin were under threat, although that soon subsided to a long period where the ALP were tipped to win Hasluck, Stirling and Kalgoorlie, while going close in Canning and Forrest.
STATE ONLY POLLS
But with WA the state polls is where it's at. It's worth noting here that while the Coalition have maintained a lead over Labor for most of the year the swing has been to the ALP. Thus for most of the year the ALP have been in front in Hasluck and Stirling, with an occasional lead in Kalgoorlie. However the Coalition have performed well in WA in November, with two polls suggesting they won't lose a seat and one of those putting them in front across the state. This lead, the first for the Coalition in any state all year, would give the Coalition the ALP held seats of Cowan and Swan.
VERDICT: Some quite different results in polling, combined with the fact that the Labor member for Cowan has retired, makes these seats hard to pick. The polls suggest Labor are more likely to make gains than the Coalition, but I wouldn't be surprised if the parties swap a marginal and the status-quo remains.
TASMANIA
ALL POLLS
STATE ONLY POLLS
As is clearly shown, a sea of red in both the national and statewide polling. While some of the statewide polling has put the TPP swing to Labor only a few per cent beyond what would be necessary to take back Bass and Braddon the unanimity at state and federal level makes it look clear-cut.
VERDICT: All statewide and nationwide opinion polls strongly suggest the ALP will regain the two Tasmania seats they lost in 2004, although the limited number of statewide polls does mean a rough chance of the Liberals to hang on to one of their two seats. Five out of five for Labor in Tasmania.
THE TERRITORIES
Both Canberra seats are safely Labor and as such all federal polling has shown the ALP extending their already comfortable margin. The only state level polling covered here is Newspoll, which includes the ACT with NSW in its cumulative statewide analysis. Once again, all examples would give the ALP comfortable control of each seat.
VERDICT: Two easy wins for the ALP.
The Northern Territory is generally not included in major polling, and therefore the polling really can't tell us anything. The only possible indicator would be the national swing, which has been comfortably enough for the ALP to win Solomon from the CLP. By the same method, Lingiari would stay comfortably Labor.
VERDICT: The federal poll says both seats for Labor but it would be a brave person who would follow that unquestioningly. Based purely on polls I'll have to say 1 or 2 seats for Labor.
OVERALL VERDICT
No prediction that is made and is based predominately on the opinion polls for 2007 could fail to give the ALP government. Adding up the bottom-end predictions for Labor listed above would give them an overall gain of 18 seats, which would be enough to govern in their own right. Adding up the top-end predictions made above would give Labor 34 new seats and a total of 94 seats. The one spanner in the works may be that there have been no new polls conducted wholly after the ALP's launch last Wednesday. If this fails with the voters we may see some polls which have the Coalition in a position where victory is possible. But if not they only have three hopes: that they can find 4 or 5% fast, that they have pulled off an unprecedented marginal seat campaign enabling them to win with well under 50% of the vote, or that twelve months of polling is wrong.
Due to the small number of ALP seats in Queensland, and the high Coalition TPP vote in 2004 in the state, the graphs for Queensland assume the safety of three ALP and six coalition seats. This leaves 3 ALP and 16 Coalition seats being analysed.
ALL POLLS
Although results early in the year placed seats such as Fisher and Fairfax at risk, since winter most polls have suggested that five seats (Petrie, Hinkler, Flynn, Bowman and Dickson) would be the most hard-fought in the election. The ALP have always been ahead in Bonner and Moreton, while generally ahead in Blair, Herbert and Longman. While a handful of polls have placed safer Coalition seats such as Forde and Leichhardt at risk the majority would see these seats retained by the Coalition with reduced majorities.
STATE ONLY POLLS
Focusing purely on state polls provides something for everyone. Coalition supporters can highlight polls in early October and mid November which minimized their losses to within two and five seats. Both polls were conducted by Galaxy polling and limited to the marginal seats of Bonner, Herbert, Longman and Moreton. The weakness here for the Coalition is that the small swings in marginal seats may not be enough as the polling ignores other seats which would also fall to the ALP if the statewide polling is correct. Galaxy may hold promise for the Coalition in these marginals but the statewide swing would probably give Labor other seats, such as Blair, Petrie and Bowman.
Further there are also polls which look good for Labor. Every non-Galaxy poll gives the ALP at least six seats, with the very safe seats of Fisher and even Fairfax under threat if statewide swings are evenly distributed across the state. While Galaxy's most recent poll shows the Coalition losing just two seats it was conducted at the same time as a Morgan poll which suggests the ALP would gain all sixteen seats considered here.
VERDICT: Queensland is going to make at least one of Galaxy or Morgan look pretty silly. On the whole the data suggests good gains for the ALP, although the Morgan prediction is a little over-the-top. The Coalition may be able to safe Herbert, Longman, and Hinkler due to local factors but the polls suggest the ALP will gain between six and ten seats in Queensland.
SOUTH AUSTRALIA
ALL POLLS
The graph of the year's national and state polls shows South Australia to be very stable in terms of seats won. Every poll has the ALP favoured to hold their marginals of Adelaide and Hindmarsh while also winning Kingston, Wakefield and Makin from the Liberals. Beyond this, most polls have the ALP winning Boothby and Sturt, although an increasing number have Sturt as too-close-to-call. While a few early polls gave the ALP a chance in Mayo and Grey, only one poll in the campaign moved these out of the Coalition's total.
STATE ONLY POLLS
The statewide figures are one of steady improvement for the Coalition, although at no stage have they come close to holding one of their three marginals and only twice have they been within 1% at Boothby. While in March the state was the Coalition's weakest, with the cumulative Newspoll giving the ALP all but Barker, recent surveys have seen Mayo and Grey return to the Coalition and their position in Sturt improve slowly. The recent Galaxy poll was the first statewide poll, and only the fourth overall, to give the Coalition over 51% in Sturt.
VERDICT: The polls have SA as a very strong state for Labor. Kingston, Wakefield and Makin would require the mother of all local campaigns to not fall to Labor, while Boothby sees Labor favoured and Sturt has them no worse than par with the Liberals. Both seats may be decided by the merits of their respective candidates (in Boothby the Labor candidate has had her share of troubles, while in Sturt the Labor candidate has been touted as a future star but is up against long-term member and Government Minister Christopher Pyne) but it wouldn't be a surprise to see Labor win both. But the only possible analysis of the polls is a three to five seat gain to Labor, with Boothby leaning Labor and Sturt leaning Liberal.
WESTERN AUSTRALIA
ALL POLLS
Due to the large difference between the federal swing to Rudd and the swing recorded in state polls, WA is the one state where looking at the federal graphs is not so helpful. For what it's worth, early polls indicated that Coalition strongholds such as Pearce and Curtin were under threat, although that soon subsided to a long period where the ALP were tipped to win Hasluck, Stirling and Kalgoorlie, while going close in Canning and Forrest.
STATE ONLY POLLS
But with WA the state polls is where it's at. It's worth noting here that while the Coalition have maintained a lead over Labor for most of the year the swing has been to the ALP. Thus for most of the year the ALP have been in front in Hasluck and Stirling, with an occasional lead in Kalgoorlie. However the Coalition have performed well in WA in November, with two polls suggesting they won't lose a seat and one of those putting them in front across the state. This lead, the first for the Coalition in any state all year, would give the Coalition the ALP held seats of Cowan and Swan.
VERDICT: Some quite different results in polling, combined with the fact that the Labor member for Cowan has retired, makes these seats hard to pick. The polls suggest Labor are more likely to make gains than the Coalition, but I wouldn't be surprised if the parties swap a marginal and the status-quo remains.
TASMANIA
ALL POLLS
STATE ONLY POLLS
As is clearly shown, a sea of red in both the national and statewide polling. While some of the statewide polling has put the TPP swing to Labor only a few per cent beyond what would be necessary to take back Bass and Braddon the unanimity at state and federal level makes it look clear-cut.
VERDICT: All statewide and nationwide opinion polls strongly suggest the ALP will regain the two Tasmania seats they lost in 2004, although the limited number of statewide polls does mean a rough chance of the Liberals to hang on to one of their two seats. Five out of five for Labor in Tasmania.
THE TERRITORIES
Both Canberra seats are safely Labor and as such all federal polling has shown the ALP extending their already comfortable margin. The only state level polling covered here is Newspoll, which includes the ACT with NSW in its cumulative statewide analysis. Once again, all examples would give the ALP comfortable control of each seat.
VERDICT: Two easy wins for the ALP.
The Northern Territory is generally not included in major polling, and therefore the polling really can't tell us anything. The only possible indicator would be the national swing, which has been comfortably enough for the ALP to win Solomon from the CLP. By the same method, Lingiari would stay comfortably Labor.
VERDICT: The federal poll says both seats for Labor but it would be a brave person who would follow that unquestioningly. Based purely on polls I'll have to say 1 or 2 seats for Labor.
OVERALL VERDICT
No prediction that is made and is based predominately on the opinion polls for 2007 could fail to give the ALP government. Adding up the bottom-end predictions for Labor listed above would give them an overall gain of 18 seats, which would be enough to govern in their own right. Adding up the top-end predictions made above would give Labor 34 new seats and a total of 94 seats. The one spanner in the works may be that there have been no new polls conducted wholly after the ALP's launch last Wednesday. If this fails with the voters we may see some polls which have the Coalition in a position where victory is possible. But if not they only have three hopes: that they can find 4 or 5% fast, that they have pulled off an unprecedented marginal seat campaign enabling them to win with well under 50% of the vote, or that twelve months of polling is wrong.
| 72 |
| Vote |
