Newspoll MkII
Posted by Possum Comitatus on October 30, 2007
An interesting little question popped up in Newspoll, and to fully chew it over properly, it needs to be put in the appropriate context:
It looks like most people are pretty locked in on this question, with even numbers of ALP and Coalition supporters suggesting that they may cross sides in the event of a rate rise. The alternative is that maybe people don’t realise how they will act until they actually see the money coming out of their bank account.
History gives us a good suggestion here of how people actually react to rate rises as opposed to whatever they may say about how they would react.
The other thing of note on the Newspoll vote was the low minor party support. This gets us on to how the small minor party vote estimates are highly volatile because they are so small.
To highlight this, we’ll go through a two step process. First, if we subtract the TPP vote of Coalition from the TPP vote of the ALP for every Newspoll since 2006, we’ll get a TPP spread. Then if we do the same for the primary votes, we’ll get a primary vote spread. These two together show us how much of the difference in the TPP vote is explained by primary votes. If we then subtract the primary vote spread from the TPP vote spread, we end up with a figure that represents how much of the TPP spread is caused by the minor party vote.
That might sound a little complicated, but it’s pretty easy once you chart it.
If we blow that minor party residual effect up to highlight its movement we get:
What this chart tells us is how much of the TPP spread is explained by the minor party vote. This minor party vote is relatively volatile as a result of it being a small number in a larger pool. For instance, if there was a survey of 100 voters and 40 were voting for party A, 40 for party B and 10 for party C, if in the next survey it turned out that an extra person was voting for party B, the vote for party P would become 41, which is a 2.5% increase, but a 1 person increase in Party C (from 10 to 11) would be a 10% increase in the Vote of Party C.
With the sizes of the polling samples we use in Australia, this effect never fully washes out of the system. So small party votes estimates are, proportionally to their vote size, more affected by sampling noise than the major party votes.
As a result of that, and because TPP estimates are based on the distribution of those noisy vote levels, a fair bit of the TPP spread is quite noisy. Hence the TPP estimates bounce around more than they would be moving in reality.
Possums Pollytics 
kwoff.com said
Newspoll MkII « Possums Pollytics
An interesting little question popped up in Newspoll, and to fully chew it over properly, it needs to be put in the appropriate context.It looks like most people are pretty locked in on this question, with even numbers of ALP and Coalition supporters s…
Pi said
Great analysis.
So it goes right back to what everyone has been saying all along. It’s not necessarily the 2PP that people should be watching, but the ALP Primary. And that has been consistent at around 48-50 all year. A very high entrenched Primary is almost impossible to break down, unless the leader implodes.
George said
I totally missed that Poss – makes total sense, and explains the TPP movement. But as you and most everyone else has been saying, the Labor primary vote is quite stable regardless of what happens to the TPP.
Also, I read in another comment here that this newspoll was taken in predominantly safe coalition eats – any truth in this?
George said
Actually I think it was more a question posed rather than a statement on the safe coalition seats – and that should “seats” not “eats” 🙂
John V K said
Yes.
Just like in opinion or media, for the levels of support minor parties get, they recieve a far bigger grab in the media. Greens in particular.
The trend line may also be showing they are being marginalised in the electorate, this time around.
The greens key issue, the environment is now major party domain with it being and ALP strength. So Bob Brown probably had no choice in prefs anyway and why he put no demands on the table except for senate prefs.
rabitoh said
I initially read that comment as a statement too George, but on a re-read it’s a question.
Doug said
Nielsen and Morgani in recent polls have shown Greens tracking around the same level of support as last election around 8-9%.
Newspoll if my memory serves me correctly has had something of a history asof underestimating the Green vote when measured against electoral outcomes.
can anyone confirm this/
John V K said
The lag effect is prob a coalition negative and not a labor positive.
Labor can’t influence rate decisions.
SwingingVoter said
I seems to me the lag in change in cash rate to lower coalition TPP votes is at least 1 quater (3 months) – perhaps reflecting the time it takes for a rise in mortgage rates to start hurting the hip pocket.
If this is the case, then based on the entrenched views in the Newspoll survey, an interest rate rise during the campaign, all else being equall, could have little impact on votes.
However, in the last few months many lenders have been raising rates 20-40 basis points without there being a corresponding RBA cash rate rise. So with these recent mortgage rate rises perhaps another RBA rate rise during the campaign may create the adverse rates perception which will impcat the coalition’s TPP vote after all.
I think I just confused myself…
GS said
George – ‘safe coalition eats’ would be caviar and fois gras? sorry:)
Possum Comitatus said
The Green vote does seem to be downplayed in Newspoll, but unfortunately we’ll never know until the election results come in.
SwingingVoter – It’s interesting whether an interest rise will affect voters in its usual way, or whether the effect will come early because of the intense media focus happening with it due to the election, or whether most voters have already built that interest rate effect into their current vote.
All three are possible, combinations of all three are possible!
Helveticus said
Possum, thanks for your relevant stats and critical analyses.
Appart from the three possiblities and combinations you mentioned, interest rate focus might also firm up the undecided voters sooner, and/or reduce minor party vote, resulting in higher primary for the major parties. Hence, it may suit both majors to battle it out over interest rates.
It's time said
Possum
I’ll add a complication to your very good statistical analysis. The major banks are already threatening to raise their rates because of the fallout from sub-prime mortgage failures. This may impact on Australian mortgage holders independently of any Reserve Bank decision. Also the general public has been primed to the prospect of a rise in mortgage costs and, with an election imminent, be quicker to connect those additional costs and their political response.
So, in the near future, I think that the lag time will be much shorter, probably so short that it will show up in a poll before the Federal election.
bryce said
Not on topic, but is poll related. I posted this elsewhere on the Morgan polls of last week (26 Oct). It relates to the table showing results for Tasmania derived from the previous four Morgan polls.
A close look at the Morgan site reveals (what seems to me) some pretty dodgy stuff. This Tas poll was the result of the Tas component of the last two F2F and last two phone polls. Total sample, without looking them up, about 1100+1100+600+600 to be generous.
For the Tas group to have any significance would mean a sample of at least 300. This would be a moe of around 5.5% Pretty large – but this is for a two horse race. When the Greens, with a solid number like 20%, are introduced the moe has much less meaning.
Now to top this off…
The total sample of these four polls is 3400. If the Tas component was sampled according to population state by state then the sample would be about 90. WTF is going on here. Either the Tas poll is 90 and the result is total crap or the last four Morgans have deliberately over-sampled Tas by more than 300%!!! This would make a joke out of the last four Morgans when it came to the result for the nation.
Have I interpreted this correctly or is there a reasonable explanation?
Leopold said
I can inform those who are interested that Newspoll’s last 4 surveys had the Greens bang on target at 7 percent in 2004.
However, that was up slightly on the position in previous months which had been 5-6. 5-6 would be my guess for Greens in 2007… unless Kevvie gets rumbled with a chainsaw up his jumper and a sack of coal down his trousers.
Although some might argue that’s already happened… 😉
KC said
Poss
Good point re the minors and how a it only needs a small change in a poll to produce a large swing in the result for them.
I think greens around the 8% in the HoR but around 11-12% in the Senate.
Labors primary is high because people are more determined this time to not repeat 98, but they want some check in the senate.
On interest rate increases, this next one will hurt so close to christmas, it affects house loans as well as personal loans and creidt cards.
And people are not dumb, since 2002 interest rates have moved up each time, there has been no decrease, in fact we will have had 10 increases on the trot. This has never happened before.
canberra boy said
bryce (#14) no there does not appear to be a reasonable explanation – in fact Gary Morgan should probably issue an apology to the polling profession and stop quoting Possum’s blog in support of Morgan’s accuracy.
I’ve gone back to check the sample numbers claimed by Morgan. There were 4580 respondents nationally to the F2F polls taken during September. Assuming the Tas sample was in proportion to enrolled voter proportions, that gives 117 voters in the Tas sample, and an error margin of 4.9% on the Greens’ estimated 8% vote and a 9% error margin on the Labor estimated vote of 56% for Sept.
Turning to October, the three F2F polls had a sample of 2927 and the two phone polls a sample of 1118: a total of 4045 nationally. Again assuming strict proportionality, that would imply a sample of only 103 for Tasmania! The 20.5% vote estimate for the Greens would be subject to a 7.8% error margin, while the 38% for Labor has an error of +/- almost 9.4%.
The poll results are consistent with the possibility that the underlying support in the community in both periods was in the vicinity of 12 or 13% for the Greens and about 47% for Labor and did not change. It was apparent long before the announcement that the Government was going to give the go-ahead to the Gunns mill and that Labor would support the decision, so I think that Gary Morgan’s ‘explanation’ for the dramatic poll turnaround is overblown.
Well spotted, bryce!!
Verity said
It is an interesting question as to how much the interest rate question has affected voting intentions.
Something that hasn’t been mentioned in all of the talk of polls is that last week Howard announced a pork barrel for pensioners and self funded retirees. This is far more likely to be the reason for the small change we see as a few pensioners are more interested in their short term future and their ability to pay the bills than the long term worry over interest rates.
It would be interesting to know the demographic of those whose votes have changed.
Robert Beswick said
Verity @ 18. That’s an good point. It might explain the movement from “others” to the Coalition primary vote. Some older folks may have been moved by the annoucement to change from FF or conservative independents to the giving a firmer nod to the coalition. Once again Newpoll’s preferences distribution formula is the wild card.
Mark said
Howard already owns the over 65s demographic so the impact of the pensioner bribe was probably pretty small. Presume many of you saw the 7.30 Report last night and the doddery old couple interviewed as typical of his Bennelong base. Contrasted with a street dancing Maxine and the smooth young Asian-faced accountant type saying “Kevin07, all the way” the imagery was pretty powerful. I am now writing to Janet Albrechtsen to demand an inquiry into the obvious conspiracy between the ABC and the nefarious Worm.
Chatswood Statsman said
Interesting to see Chinese and Korean Australians finally expressing their hurt at JWH’s support of Pauline Hanson. Ironic that the electorate of the PM has literally shifted under his feet to include so many people of colour.
Pi said
Bryce and Canberra boy…
I believe Morgan is over-sampling Tasmania, and using a weighting formula for the rest of the country.
This leads to the obvious question. What is really happening in the big population centres of Melbourne and Sydney, specifically places like Higgins, North Sydney, Kooyong, and Bennelong?
palooka said
I don’t know about you, but when I do an Australian poll, I always put in that I come from Bennelong.
alpal said
Possum,
I’m led to believe – and am easily led – that Goldstein is very much in play. I also think the warm and caring Danna Vale is in strife in Hughes. These seats are outside the notional national numbers. Is it the demographics, the candidates or the mood – or all three?
Possum Comitatus said
Helveticus at 12,
That’s a real likelihood, especially for those that have moved across to the ALP on the basis of interest rates already.
It’s Time at 13,
Not only might the effect come through quicker for that reason, but it might have already been priced into the vote with all the attention given to both the banks lifting rates unilaterally and the likelihood of a rate rise.
Newspoll might produce a demographic breakdown before the election, so we might be able to pick up the over 50s if they actually moved over the last 2 weeks.
On the Tassie/Morgan thing, thatnks for that folks, I’d never actually thought about it.
Sounds all a bit dodgy whichever way it cuts.
Alpal,
There’s been a lot of action in Goldstein, there’s been polling done there which is just unheard of, and Andrew Robb looks grumpier than usual. The idea of Goldstein voting Labor actually makes me laugh, but it looks like there is a serious swing on in those toffy seats like Goldstein, North Sydney and those aspirationally toffy seats like Aston and Casey. Go figure.
I think those small L liberal seats that are moving for the multitude of reasons I outlined at
and
https://possumcomitatus.wordpress.com/the-bangkok-dilemma/
Which will make it hard for Howard to keep both the leafy seats, and the regional seats – as the more he plays to one, the more he alienates the other. Demographics and mood would be my bet.
Stig said
With the MoE noise around the minor vote, it doesn’t look like there’s a lot of change there over time. It will be interesting to see the results in a months time though.
I have to withdraw a previous whinge about MoE non-reporting – thank you Kerry O’Brien and Antony Green for being out there. There are still plenty of culpable morons out in the MSM, but I don’t have to listen to them.
Is it just me, or is this campaign dragging a bit? As a self-confessed political junkie who has been hanging for the campaign all year it’s come on quite flat so far. I’m still enjoying it, but it just seems like Howard’s got nothing to say and that he’s ducking the media a lot. Just sayin…
El Nino said
Possum, this is a little off topic but… To what extent do you think that interest rate rises contribute to wages growth? A tight labour market combined with highly geared mortgagees and a strong consumerist ethic could conceivably combine to cause interest rate rises to have some upward pressure on inflation.
I am not saying that the net effect of interest rate rises is upward pressure on inflation, but under the current circumstances it may be that rate rises are less effective in combating inflation.
I am unschooled in economics so I might have just stumbled onto something that someone else cottoned onto 100 years ago, but that is the nature of blogs…
From a pseph point of view I am looking for an angle where interest rate rises have an indirect (and maybe delayed) influence on voting intentions.
Samuel K said
I recently watched a mini-doc on the Australian website showing how the Newspoll is brought to us each week. It showed what I consider to be two potential causes of bias in the Newspoll process:
1. Nearly all of the people asking the polling questions were younger people, mostly university students
2. The questioners were reporting a significant number of knockbacks – e.g. taking six separate calls before someone is willing to take the time to respond to the poll
For 1, is it possible that people subconsciously prefer to please the interviewer by saying Labor because they know that younger people are strongly aligned to the Labor party (particularly for this election)? – And assuming that one can tell that the questioner on the other end is young.
I would consider this bias to be similar to what I believe is an inherent deficiency in the Morgan face to face polling. Some people feel slightly guilty about saying that they will vote Liberal since it is seen as the individualist’s party, the party of money and business and not social issues etc. This may be part of the reason for the Morgan’s consistent pro-Labor bias.
For 2, and this is a bit more far-fetched, is it feasible that more conservative types are grumpier/consider themselves shorter of time and are less likely to provide a response?
Possum, I would be interested in your thoughts on these potential sources of ‘bias’.
ismark said
My experience has been that young people show the diversity of opinion that older people do.
It is oft said that the sins of the father are visited on the child. I certainly believe that political affiliatiion is a matter of nuture not nature. So as we on this site differ in our opinions so do our children.
I think it is erroneous to say that people, when questioned by a younger person, will assume that they vote labour and accordingly feel the need to give them the answer that they feel they want.
Having just had to discuss “how to vote” with my 18 year old daughter it is an adventure akin to explaining the birds and the bees.
ismark said
Thinking about it further,
respondents who take the time to answer the phone at dinner time would never be “Dear old Ducks” with nothing better to do or people who are far to polite to say “Bull Shit” who might vote conservative would they.
Maybe the bias balances or maybe all polls are to be taken in perspective as a rough indicator of the mood of the population and not a definitive indicator.
Enemy Combatant said
I still think Ryan is gone, more so everyday.
Comment by Possum Comitatus — October 29, 2007 @ 2:23 pm
Acting on information received I’ve just had a serve of Ross Daniels in Ryan @ $2.95. Little fish are sweet. Good oil is sweeter.
Comment by Enemy Combatant — October 29, 2007 @ 4:07 pm
48 hours later……..
“The depth of the Liberals’ problems can be gauged by the knowledge that they have virtually written off Mr Johnson’s chances in Ryan, even though he enjoys — on paper — a splendid margin in the Liberal heartland.”
http://www.theage.com.au/news/federalelection2007news/storm-clouds-gather-for-liberal-mps-in-the-sunshine-state/2007/10/31/1193618975785.html
Come to think about it, cockroach capitalism ain’t all that bad.
mark-sydney said
So Poss, does the absence of postings from you yesterday mean we can look forward to multiple postings today : )
Withdrawals happening here….
Jack Rodios said
Hi Possum,
Great summing up of the polls, but is it credible that in reality the numbers for this election should be run on 2004.
Howard would never have got a 16 seat majority without being able to foster fear of Latham.
And in my view, there would be very few voters who switched to Howard in fear of Latham, that won’t automatically drift back to Labour ander the seemingly safe hands of Rudd.
This would mean that the margin Rudd needs is more like 2% and subsequently make Howards chances even slimmer.
Possum Comitatus said
There’s probably alot of truth in that Jack, but the problem becomes trying to figure out that 2004 election effect and apply it to today.
The swing at the moment that the polls are suggesting probably contains the two components of the readjustment from the last election as well as the actual swing that Rudd has generated over and above that. Trying to actually quantify those two swing components though becomes pretty difficult.
If Labor wins, after the election we might be able to do it using some panel data regressions, but until we actually have those results in, all we can do is speculate.
All part of the fun really 😉