Report on Oz here. This is a huge change in a single poll; see from the tables that preferred PM numbers remain as lopsided as always, an odd state of affairs.
As always with such movements, we have to wait for the next polls to see if this means much. Later in the week Morgan will hopefully publish one that was taken over the last weekend. Essential Research already has, and they still have Labor sky-high. But their polls are online, tend to favour Labor and use a possibly unrepresentative sample. Still, you would have expected some movement at least in Essential if Newspoll is ‘accurate’.
So a waiting game. Depleted swagger quotient in the PM’s office this morning. What is internal Labor polling saying? Will they panic and do something silly?
This represents a 15 point turnaround in the primary vote gap, ie from a lead of 15 in the last poll to zero in this one.
This table below shows all such Newspoll primary vote turnarounds over 10 points since they began in the mid 1980s. The entries in purple italics are the big movements, and I’ve included Newspolls immediately before and after. The change column number refers to Labor, ie positive is to Labor and negative to Coalition.
Some are easily attributable: Andrew Peacock’s elevation to the Liberal leadership in May 1989, the famous 1993 Dawkins budget (the biggest movement of all), the volatility of Alexander Downer’s 1994 leadership, Howard becoming leader in Jan/Feb 1995, September 2001 (although why such a big jump in that single poll is not clear update: Sept 11, duh). [Update: October 2002 is of course Bali bombs. Simon Crean was actually ahead after preferences before that.]
Some were corrected to an extent at the next outing (Sept 1986 and Aug 1995) and some were themselves in part corrections for previous poll (May ‘89 again, and there’s a wee touch of that in today’s Newspoll, as last fortnight’s moved to Labor by 4 percent).
Stay tuned for next Newspoll and others.
Newspoll Date |
Labor primary |
Coalition primary |
Change in primary vote gap from previous poll |
| 22-24 August 1986 | 46 | 43 | 3 |
| 19-21 September 1986 | 40 | 50 | -13 |
| 17-19 October 1986 | 43 | 47 | 6 |
| 5-7 May 1989 | 49 | 41 | 9 |
| 19-21 May 1989 | 40 | 45 | -13 |
| 2-4 & 16-18 June 1989 | 39 | 45 | -1 |
| 23-25 October 1992 | 37 | 49 | -3 |
| 6-8 November 1992 | 45 | 41 | 16 |
| 20-22 November 1992 | 44 | 43 | -3 |
| 6-8 August 1993 | 41 | 46 | -3 |
| 20-22 August 1993 | 31 | 54 | -18 |
| 3-5 September 1993 | 33 | 50 | 6 |
| 13-15 May 1994 | 46 | 44 | 3 |
| 27-29 May 1994 | 42 | 51 | -11 |
| 10-12 June 1994 | 41 | 49 | 1 |
| 19-21 August 1994 | 41 | 48 | -6 |
| 2-4 September 1994 | 46 | 42 | 11 |
| 16-18 September 1994 | 50 | 38 | 8 |
| 23-25 September 1994 | 42 | 46 | -16 |
| 7-9 October 1994 | 44 | 44 | 4 |
| 27-29 January 1995 | 46 | 40 | 8 |
| 10-12 February 1995 | 38 | 47 | -15 |
| 24-26 February 1995 | 41 | 46 | 4 |
| 28-30 July 1995 | 38 | 49 | -4 |
| 11-13 August 1995 | 43 | 43 | 11 |
| 25-27 August 1995 | 39 | 49 | -10 |
| 24-26 May 1996 | 36 | 55 | 0 |
| 14-16 June 1996 | 40 | 48 | 11 |
| 28-30 June 1996 | 40 | 47 | 1 |
| 16-18 October 1998 | 39 | 47 | -8.6 |
| 30 Oct – 1 Nov 1998 | 43 | 40 | 11 |
| 13-15 November 1998 | 40 | 45 | -8 |
| 7-9 September 2001 | 40 | 44 | 2 |
| 21-23 September 2001 | 35 | 50 | -11 |
| 5-7 October 2001 | 35 | 50 | 0 |
| 27-29 September 2002 | 39 | 41 | 4 |
| 18-20 October 2002 | 34 | 47 | -11 |
| 1-3 November 2002 | 34 | 46 | 1 |
| 14-16 May 2004 | 44 | 41 | 3 |
| 28-30 May 2004 | 37 | 47 | -13 |
| 18-20 June 2004 | 43 | 43 | 10 |
| 10-12 March 2006 | 35 | 45 | -8 |
| 24-26 March 2006 | 42 | 41 | 11 |
| 7-9 April 2006 | 39 | 41 | -3 |
I presume you may have forgotten September 11 in the 2001 poll change? This would have been polled in the aftermath of that event.
Anyhow, I’m sure we’re all looking forward to the “We’re Back” posturing over the next fortnight.
You’re right! I kind of missed the fact this that was the next poll after that quite well-publicised event.
The September 2001 polls are interesting. The Tampa Affair had largely happened by the time the 7-9th Sept 2001 polls was done; the legalisation removing big chunks of Australia from our migration zone passed the week before. So it appears that it was not as massively popular as popular history would have us believe.
Then the 9-11 attacks happened and the doo-doo hit the fan so to speak. Every government around the world suddenly became extremely popular. Bush got approval ratings of 90%+ (and there was some relief when Howard got back to Australia from the USA… largely because Downer and Costello had been acting PMs).
The next poll you give is at the start of October which was when the war in Afghanistan was breaking out. So there is some of the same flag waving sentiment in that poll too.
But within a month of that poll the election result was 51-49 which is a lot closer than the polls would suggest.
I really wonder about the accuracy of the poll. Why is it that only the ALP & Coalition primary votes shifted dramatically? There was no change in the Greens & “Other” primary votes, suggesting the voters changing to the coalition were all hardliners on asylum seekers, even though Rudd copped a lot of personal criticism over being too hardline or being too weak, depending on your perspective. Surely there would have been some shift to the Greens & a correlating dramatic shift downwards in Rudd’s performance figures & preferred PM stakes. This smells a bit to me. Noone should forget that the OZ has agendas to follow.
shades of the “missing newspoll”
JA,I think we can assume there is no way Newspoll would do anything like that, even if asked.
I share John Anderson’s incredulity about the static support for the Greens in this poll. Surely there are a number of disenchanted Labor voters out there who would be siding with the Greens at the moment over immigration policy. Perhaps the lack of change is due to the fact that the Greens’ primary vote has already been quite high at around 10% for the last few polls. Perhaps the kind of voters disenchanted with the Rudd Government on immigration have already been disenchanted about the Government’s stance on the ETS and that might mean we are unlikely to see any further drift from Labor to the Greens. I guess we’ll have to see what trend emerges over the next few fortnights.
Many years ago, when I was at Uni, I did one of my chemistry labs with a guy who worked for one of the pollsters. This is in the late 80s/early 90s. Anyway, the pollster had been market testing one of the Lib promos at the time, it might have been “It’s Time for Plain Thinking” or maybe “Incentivation” or something else. Anyway, this character personally didn’t think much of the package, and he told me that he was misrecording results to make it seem like the package was less popular than it really was.
I suggested to him he might get into trouble, but he said they would have no way of knowing. I was not sure if he was right or wrong, although it occurred to me that clever use of statistics might pick him up.
I would assume in this day and age these calls are recorded, so it’s a much more risky proposition for those who staff the pollster call centres to record inaccurate results. I would assume that the organisations do at least some checking (for “quality and training purposes”, as they say), particularly around a rogue poll. I think it is unlikely however, that they check that the responses every person polled was properly recorded, even in the context of a rogue poll.