Home | site-map | indonesia | indonesian-single-source-articles | asteroid-user-registration |
Roy Morgan OZ Panel Login | Mail Diary Panel Login | careers | contact-us | state-of-the-nation-webcast |
|
||||
 |
 |
 |
|
  |
|
|
||||
COMPANY ONLINE STORE PRODUCTS SERVICES INDUSTRIES MORGAN POLL PAPERS PRESS RELEASES CONSUMER CONFIDENCE READERSHIP UNEMPLOYMENT THE REACTOR 
|
||||
| NEWS : Morgan Poll : | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Galaxy Poll Most Accurate - Morgan Poll Most Accurate on Major Parties, but Overestimated Greens and the Flow of Preferences to Labor
The most accurate poll in the 2004 Federal election was the telephone Galaxy Poll conducted October 5/6. Their average error for the major parties was (0.6) and for all parties (0.4). They predicted the significant swing to the L-NP. The telephone Morgan Poll was equally accurate in predicting the swing to the L-NP and the vote for the major parties (the average error of 0.6). Although the telephone Morgan Poll accurately recorded a 16% vote for the non-major parties as a whole (the actual other vote was 15.3%), it did not accurately represent the breakdown of the ‘other’ vote — specifically the Morgan Poll over-estimated the Green vote at 9.5% (actually received only 7%) and underestimated the vote for the new conservative Family First Party, which actually received 2% of the vote ahead of Australian Democrats and One Nation (1.2% and 1.1%). While Newspoll, like the telephone Morgan Poll, wrongly said the election would be close (Newspoll a small L-NP win, Morgan Poll “too close to call”), Newspoll was less accurate than the Morgan Poll in predicting the major parties (average error, 1.1 for Newspoll) but more accurate when the minor party errors were taken into account (Newspoll’s average error for all parties was 0.6). We have always been of the view that face-to-face polling is the more stable and accurate polling method, however, face-to-face polling cannot be conducted during the days before the election. The last pre-election face-to-face poll is conducted the weekend before the election, and we have seen once again that a lot can happen in a week, and the vote can change. - Tasmanian CFMEU divided the union movement by supporting the Howard L-NP Government on their policy for continued logging of Tasmanian old growth forests. This polarised the electorate. - Green’s leader Bob Brown came out strongly and emotionally in support of the ALP against the L-NP, not only in relation to the logging of old growth forests in Tasmania — but in relation to many other issues. The electorate was further polarised. Thus, Bob Brown opened the door for the Liberals to attack the Greens’ proposed ‘loopy laws’ and the ALP/Greens preference deal that could have handed Government to Labor. It also created what would be called in normal marketing a “market gap” — the Greens no longer represented an ‘anti-major party’ vote (the Family First Party was ready in the wings to fill that gap — and did). This means we must use telephone polls or another alternative, to conduct interviews during the week before the election (The Bulletin/Channel 9 showed that Internet polling is still not reliable). Although television advertising came to a halt on the Wednesday before the election, newspaper advertising continued, and telemarketing began (in another first for Australia , the Prime Minister, in the days before the election, undertook a major telemarketing exercise), and ‘point-of-sale’ (voting booth) advertising was in full force, as it was at the 2001 Federal election. It was always going to be difficult to measure and understand the electorate in the lead-up to this election — with many difficult issues pulling segments of the electorate in different directions and strong anti-major party feelings on such issues as the War in Iraq, treatment of asylum seekers and Tasmanian old growth forests. We were concerned that the electorate’s stated support for the Greens was higher than would be reflected in the actual vote. It appeared as if the Greens had become ‘code’ for ‘not ALP or L-NP’. But the real ‘sleeper’ was the Family First Party. This church-based party was a well kept secret — right up until a week before the election. Family First did not build awareness, interest in its policies or support for its position. Rather, all effort was directed toward exhaustive and active manning of the polling booths. Family First picked up 2% of the national vote and more than 4% in some seats — higher than the Australian Democrats and One Nation. It should also be noted that both the Morgan Poll and ACNeilsen predicted a week before the Senate election, a 12% vote for the Greens in the Senate — they in fact only received 7.5%, and Family First is likely to gain a Victorian Senator and 1.8% of the national vote. It now looks as though the L-NP can control the Senate with them - thanks to the Victorian ALP preferencing Family First before the Greens!
For further information: Gary Morgan : Office (03) 9224 5213 Mobile 0411 129 094 Home (03) 9419 3242 Michele Levine : Office (03) 9224 5215 Mobile 0411 129 093 Home (03) 9817 3066
#Australian Electoral Commision 12.00pm October 10, 2004.
#Australian Electoral Commision 12.00pm October 10, 2004. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
© 2007 Roy Morgan Research. All Rights Reserved |
