There were a number of key moments in the last days of the campaign which we specifically asked about in the Exit Poll. We also tried to capture the impact of Bob Hawke’s death across both the Exit Poll and Final Reflection surveys.
The specific key events tested were primarily the set pieces: the two party launches and the last debate. The Labor Party Launch had the most impact in terms of switching vote, although it should be noted that few of those were ‘converted’ from right to left, they were mostly moving from minor parties to the ALP. The Liberal Party Launch by contrast caused almost no one to switch votes, but performed better in reinforcing the party’s vote. The debate caused literally no one in our panel to change their vote intention. This really was preaching to the choir, with a strong confirmation of vote number and the vast majority saying it had no impact on their vote.The other key moment we asked about was Bill Shorten’s appearance on Q&A.And the subsequent story about his mother in the Daily Telegraph, and his teary, defiant response, which was hailed by many as a decisive moment.It wasn’t. While there was plenty of vitriol in the comments as people vented what they thought of the Telegraph piece, neither the story, not his response to it, moved significant votes.
The last unexpected moment was asked about in the Final Reflection survey, as well as encouraging respondents to note their thoughts in the ‘final say’ open text box at the end of Weekly Survey 5. (As the survey was already in the field prior to Hawke’s death, it was not possible to change it.)These numbers should be taken with a grain of salt, as there is a great deal of post-rationalisation and the bias of distance in all the final reflection responses (which calls into question the validity of all post-election surveys, but that’s a debate for a different time and place).
We know from the comments in Weekly Survey 5, completed as people were voting, that people were stirred up, moved, reconsidering, re-assessing Bill Shorten and finding him lacking when compared to Hawke. Many of our respondents also believed Hawke’s death would lock in a landslide or even cause a bigger swing to Labor. In a fairly slow process of tracking back through those voters who completed the Hawke question set from at least 3 surveys back, about half (excluding pre-poll voters) wobbled to some degree in their election day vote. That is, they changed their first preference on election day (but stayed on the same side of politics – and many reverted to their earlier vote preference: thus wobbling, not really swinging) or changed their preferences. We absolutely cannot say it was because of Hawkes’ death – as you can see above, most said it had no impact on their vote, when asked with the benefit of a fortnight’s distance.
The only way to answer this puzzle is to run the study again. If the same ‘wobbling’ is seen in the last weeks, then we know it is normal election behaviour in Australia, and not specific to Hawke’s death.