The Pollsters and the Pollies: Polling for Political and Pecuniary Profit – With Particular Reference to Newspoll

9

April 20, 2013 by kate ahearne

If you’re a big player in the media with self-interest to protect, and you’re not entirely scrupulous about how you achieve your ends, one of the most powerful tools in your shed must be the opinion poll.

During this last fortnight, there has been a flurry of polls – not necessarily coming up with compatible results, or with results that seem to reflect the actual political news as it has unfolded – some of these polls have been counter-intuitive, shall we say.

There are several proven ways you can go about using opinion polls to mislead the public, and to some extent, to lead even the pollies by the nose.

• Conduct and publish your own polls.  Newspoll is one of these.
• Pay someone to conduct polls for your exclusive use.  Galaxy is one of these for the Murdoch Press, while Nielsen has a similar relationship with Fairfax.
• Pay close attention to the methodology.
• If you still don’t get what you want, spin the results.
• If the poll has been conducted by a source over which you have no particular control: spin, ignore, or cast doubt on the methodology and/or the honesty of the results.  (This will distract attention from your own methodology, and give you a perfect alibi in case you want to accuse someone else of doing exactly what you’re actually doing yourself  You might, for instance, want to trash Gary Morgan – but I’m getting ahead of myself.)

So let’s take a closer look at the polls, with particular attention to Newspoll and the art and logic of Methodology.

Results of Newspoll’s regular ‘voter intentions’ appear exclusively in Mr Murdoch’s flagship newspaper, The Australian.  Their self-assessment goes like this:

You probably know that Newspoll is Australia’s leading public opinion polling company.   Established in 1985, we have Australia’s best track record having estimated the outcomes of every state and federal election since our company was founded.’  (From their website)

(Although, as one wag on Twitter remarked – if NewsPoll is so accurate, how did they get it so wrong in the US?  What happened to President Mitt Romney? – my ‘bold’ and my paraphrase.)

Let’s go back a bit.

On Nov 9, 2009,  Gary Morgan and Michele Levine from Roy Morgan Research cast serious doubt on Newspoll, under the headline: News Ltd’s rogue Newspoll changed Australia’s political agenda; and further inflamed the ‘boat people’ debate.

Responding to a poll by Newspoll on Nov 3, 2009, Morgan and Levne raised the question: What does it mean when a poll result is ‘surprising? (My paraphrase.)

The evidence was clear.  Yet the publication of News Ltd’s  poll (Newspoll) in the first place had already had a major impact.  The evidence showing the ‘error’ of Newspoll was literally ignored by media discussion (e.g. the Insiders on ABC TV and the impact of the ‘rogue’ poll was allowed to run unabated).

‘Last week, November 3, News Ltd’s   poll (Newspoll) claimed that Coalition support had jumped 7% to 48% while ALP   support had dropped 7% to 52% in two-party preferred terms.‘Even at face value such a drop in support was extremely   unlikely.‘On Thursday November 5, when the Roy Morgan Consumer Confidence   Rating showed an increase in Consumer Confidence (up 2.5pts to 128.0), it was clear to those who follow polls and politics that the Newspoll was wrong — a ‘rogue’ poll.  An increase in Consumer Confidence is almost invariably positive for the Government.‘On Friday November 6, the Morgan Poll showed two-party preferred support for the  ALP virtually unchanged (up 0.5% to 61%) and then today’s Nielsen poll also   showed support for the ALP virtually unchanged (down 1% to 56%).’

Mission accomplished!  The voters, and even ABC’s Insiders had been misled, swayed, and told downright lies, which most of us believed.

The public and the media had been ‘played’ like an orchestra by the maestro conductor.

But wait, there was more.  If you own the poll, you can just withhold whichever results you don’t like.  As Morgan and Levine, clearly disgusted, went on:

‘Pollsters and those who publish the polls have a responsibility  to report the facts and the truth….  It is extremely worrying that today’s   Newspoll on “boat people” clearly did include questions on ‘Political   support,’ but the results from the ‘Political support’ question were not   published.’

Methodology.

There are two extremes when it comes to methodology – you can run your polls in such a way as to produce the results you want, or you can run them so as to discover the facts.

If, for instance, you own a polling company, as in the case of Murdoch and Newspoll, or if you have an ‘exclusive’ relationship with an ‘independent’ poller, as the Murdoch papers have with Galaxy, methodology is easy- you just use the one most likely to yield the result you’re looking for.

You would, for instance, prefer telephone polling, because that would automatically give you an older, wealthier, more LNP-leaning demographic.

So you would be likely to ask something like, ‘If an election for the House of Representatives were held today — which party would receive your first preference?  If you are unsure, which would be your preference, however slight?’

You would be less likely to ask, ‘If an election for the House of Representatives were held today — which party would receive your first preference?’

And so on.

So let’s see how this worked out in the last couple of weeks.

Timeline

The week beginning April 5 …

• Wayne Swan and Bill Shorten unveil Labor’s Super policy.
• PM, Julia Gillard leaves for China.
• Mr Abbott begins a series of really silly statements, including ‘Shades of Cyprus’, and reveals his devilishly clever master-plan to rob the poor in order to leave the ‘rich perks’ alone.  We’re expected to swallow this, but most of us are not silly, so we don’t swallow it. But nobody rings us up to ask us for our opinion.
• Newspoll (April 8) shows a ‘surprisingly’ small but definite swing towards Labor.  As Crikey saw it:

The primary votes are 32% for Labor (up two), 48% for the Coalition (down two)   and 11% for the Greens (up one). Last fortnight’s spike has also come off in   the personal ratings, with Julia Gillard up two on approval to 28% and down   three on disapproval to 62%, Tony Abbott down four to 35% and up four 54%,   and Abbott’s lead as preferred prime minister down from 43-35 to 40-37.

• Ms Gillard’s week in China is a huge success.  Just one example from the press should suffice:

‘That Gillard has been able to bring relations with   Beijing back from these historic lows and announce a coup of the magnitude   this week is really nothing short of spectacular.’  Laurie Pearcey, The Conversation.

• Malcolm Turnbull reveals the Coalition’s cheap and nasty ‘two-lane Sydney Harbour Bridge’-style answer to Labor’s Broadband, to howls of derision.  (Click for Crikey)
• Simon Crean has some more really silly and/or very naughty remarks to make.

April 14

The Galaxy Poll turned up in the Murdoch paper, The Sunday Telegraph, under the headline: ‘Galaxy poll shows Tony Abbott ahead as leader voters trust on superannuation.’ 

‘Just one in three – 34 per cent of voters – agreed they were most likely to trust Julia Gillard with their super.Overall, 45 per cent of voters said they were most likely to trust Tony Abbott with managing government policy on their retirement savings.’The result is a blow to Labor’s proud record of introducing the first compulsory employer contributions to superannuation in Australia during the Keating years.Support for Labor has also flatlined, with a primary vote of just 33 per cent according to Galaxy, a result that would deliver a devastating electoral defeat if repeated on September 14.Support for the Coalition is steady at 47 per cent, the Greens on 12 per cent.Assuming a flow of preferences similar to the last election, this would deliver a two-party-preferred outcome of 54 per cent for the Coalition and 46 per cent for Labor.

Assuming a flow of preferences similar to the last election, this would deliver a two-party-preferred outcome of 54% for the Coalition and 46% for Labor.

A ‘surprising’ result, surely.

April 14

Then there was ReachTel, 7 News Melbourne

If a state election were to be held today, which of the   following would receive your first preference vote? If you are undecided to which do you even have a slight leaning?

This was a State-based Poll, without any 2-party preferred outcome, but the interesting thing is the wording of the question –  ‘If you are undecided to which do you even have a slight leaning’

April 15

The Fairfax/ Neilsen Poll was conducted 12-14 April, two days after the launch of the LNP’s alternative to the NBN.

Labor’s  national broadband network is more popular than the Coalition’s cheaper   version, according to the first national poll on the policies.Of those who   had heard about the government’s NBN, about 63 per cent of those surveyed supported it, reveals the Fairfax/Nielsen poll of 1400 Australians.However, of those who have heard about the Coalition’s alternative, only 41 per cent back it.Support for Labor’s NBN was consistently high across the states with most registering levels of support between 60 and 70 per cent.

It had been a brilliant couple of weeks for the Government, and yet Labor showed a slump in both the Galaxy and the Nielsen Poll.

In spite of the absurdity of Abbott’s insistence that he would oppose the government’s plan to tax Super earnings over $100,000 p. a. at 15% while robbing the poorest 3.6 million earners of their modest Super tax breaks; despite Ms Gillard’s triumph in China; despite the widespread thumbs-down to the LNP’s answer to the NBN – Labor’s primary vote, according to the Fairfax/Nielsen poll fell 2 points to less than 29%, while the Coalition rose 2 percentage points to 49%.  Even the Murdoch-owned Newspoll just a few days earlier, had Labor doing better.

Allowing for the relentless misinformation campaign in the media, and for the impact of Simon Crean’s atrocious behaviour, at this stage, we can only say that this result was ‘surprising’, and if Gary Morgan and Michele Levine are any guide, we ought to be suspicious. Particularly as, also on April 15, …

April 15

The Essential Poll turned up.

Q. If a Federal Election was held today to which party will   you probably give your first preference vote? If not sure, which party are you currently leaning toward?Q. If don’t know -Well which party are you currently leaning to?Sample size = 1,897 respondents

2PP	Election 21 Aug 10	4 weeks ago	2 weeks ago	Last week	This week
Total Lib/Nat	49.9%	54%	56%	56%	55%
Labor	50.1%	46%	44%	44%	45%

April 15

The Morgan Poll also turned up, not in the Mainstream Media, but on the Independent Australia site, with a linkback to the Morgan website, and the headline: Morgan poll shows Government support up for second straight week

‘Last   weekend’s Morgan Poll shows support for the LNP down to 55.5% (down 1% since   April 4-7, 2013), on a two-party preferred basis, compared with the ALP 44.5%   (up 1%).’

This was a multi-mode poll held from April 11-14 with a sample of 3,835. (Roy Morgan Research has been experimenting with ‘modes’ lately – mixing face to face with web-based polling on March 5, and more recently conducting an SMS-based poll.  Gary Morgan says that their polls from March onwards up until the election will combine face-to-face with web-based and SMS polling. So, it seems that Morgan is dragging the polling game kicking and screaming into the brave new world of mobile phones and the internet.)

People were asked:

‘If an election for the House of Representatives   were held today — which party would receive your first preference?’

Here is Morgan’s graph showing the two-party preferred vote since the end of 2009.  Not quite the unmitigated disaster for Labor that some would have us believe, and plenty of evidence of bounce, with even-pegging occurring several times during the period – most recently just 4 months ago.

(Thanks to Roy Morgan and Independent Australia for the graph.)

So, to begin with, let’s apply our magnifying glasses to that little patch in December last year – not so long ago in the scheme of things, when the two sides were neck and neck.

On Dec 17, Bob Ellis made some interesting remarks under the headline, Labor To Win: The Fairfax Poll, Decrypted

‘…this morning’s Fairfax Poll (the honest one) has Labor winning; or rather prevailing, probably, with   Green and Independent help, as it did in 2010.’

He adds another percentage point for the younger Labor voters away from the home phone in summer.  He was guessing, of course, about the specific weighting to be given to those younger voters, but some allowance was more than reasonable.

But by Feb 26, Ellis was fairly boiling.  Under the headline, The Usual Murdoch Dirty Tricks (79): Today’s Newspoll, he went to town on the fact that the poll had been weighted by Newspoll “’to reflect the population distribution’”’ but also excludes the key electoral areas affected by the floods, and, of course all those mobile phones not being rung.

‘This uncorrected omission of numbers comes under criminal fraud   I would think, and O’Shannessy should be arrested; in my view.’

Ellis’s analysis on this occasion was quite compelling (even if his emotions were not so reliable), and well worth a look at in its entirety, as is a similar, passionate but reasoned article as recently as March 30.

‘Almost no-one under fifty has landlines in Queensland and those   are all he rings…   If the Katter preferences go Labor’s way, and Katter’s detestation of Newman will ensure they will, it’s possible Labor is already   on 54 two party preferred and gaining, or Katter’s base vote is on 21 and in a position to win eleven seats and negotiate a coalition with either side.’

April 15

Also appearing on the 15th was  the Essential Report, which  had the Coalition at 55% – Labor, 45%, on a two party preferred basis.  The interesting thing to notice here, is the way the question is put – ‘If don’t know -Well which party are you currently leaning to?’

Q. If a Federal Election was held today to which party will   you probably give your first preference vote? If not sure, which party are   you currently leaning toward?Q. If don’t know -Well which party are you currently leaning   to?Sample size  1,897 respondents

Where the Morgan polls include the ‘don’t knows’ as ‘don’t knows’, Essential  follows those pollsters who ‘disappear’ them into the two-party preferred.  It’s a shame, really, that Essential, whose credentials as stated, are otherwise ‘worthy’ (until/unless proven otherwise) should stoop in this direction.

Well, that’s methodology for you!

So, of those we have looked at here, which pollster would you be more likely to trust?

1. A.  Newspoll
2. B.  Galaxy
3. C.  Nielsen
4. D.  Essential
5. E.   ReachTel
6. F.   Other

Anyway, let’s give the last word to Truth Seeker:

‘So the question remains, if you don’t have a barge pole with which to shove Abbott over the line, will an opinion poll do the job?’  His article, Abbott? when there’s no barge pole, an opinion poll will do! (posted on 12/3/13) is a doozie.  I couldn’t really find a ‘best bit’ to quote here, because all the ‘bits’ were ‘best’, so here it is:

Truth Seekers Musings tt? when there’s no barge pole, an opinion poll will do!

Posted on 12/03/2013 by Truth Seeker

Abbott? when there’s no barge pole, an opinion poll will do!

So, what’s happening with these bloody polls?

The previous two poll cycles showed a massive boost to Abbott after months of trending up for Gillard and the ALP, to almost level pegging TPP.

Now the week before the first of the poll jumps, for Abbott, he made his first major speech for the year, announcing that the LNP would remove the carbon “tax” (which is working, and the threat of removal just adds uncertainty to the business sector), wind back pensions, take away the school kids bonus and take the tax free threshold from $18,200 back down to $6,000.

Most would assume that these announcements would be less popular than a fart in an elevator among pensioners, families and in fact most tax payers… but NO!    Rather his personal popularity jumped by (from memory), around 9%.

So the question is WHY!

Vlad the impaler 9 (Photo credit: Wikipedia)

I have been living in the electorate of Fisher, a safe LNP stronghold, held for many years by Peter Slipper.

Many times I have made the point that, here in Fisher, they would elect Vlad the Impaler if he was endorsed by the LNP, as they are so rusted on.

The reason that I am relating this is because since moving here around 15 years ago, I have been polled twice in national polling and two to three times in QLD polling.   Now maybe that’s just coincidence, but I have heard many in safe Labor seats complain that they have never been polled?

So what is there that can affect the outcome of polls?

For starters, we know that these polls are taken on landlines, excluding a large percentage of the younger demographic.  Most of the younger generations that we know only use mobiles.

When you consider news papers like The Australian, or in fact any of the tabloid papers that have turned lies, spin and misrepresentation into an art form, it’s not too much of a stretch of the imagination to think that they could manipulate the polls, by the questions that they ask and the postcodes that they target. There has been much discussion about the questions asked, so lets look at the other option.

Now for instance, five minutes research on google tells me that the the electorate of Benelong, has phone numbers with the Sydney prefixes   9871    9847   9638   9427, so if you wanted to get a poll that had a strong LNP bias all you would have to do is to provide a list of prefixes that you want them to ring.

For someone like Murdoch, or that other paragon of virtue Rinehart, how hard would it be?

And just this week we see a jump of 6% in the PM’s personal ratings after two weeks of speculation about K Rudd’s comeback, even though he has denied it strenuously. And lets face it, if the preferred ALP leader, as endorsed by the rabid right, is K Rudd, and we can safely assume that they don’t have the best interests of the ALP at heart, then why are we even discussing it?

Now maybe this was simply payback for the PM taking away months of speculation about the election date and the resulting plethora of front pages that they have lost as a result?  Or maybe these last couple were rogue polls, and the latest is the inevitable correction?   Or maybe there is something more sinister going on, like an attempt at destabilisation of government through manipulation of the people via the MSM, and the polls?

Many right whingers will tell you that the influence of the polls is overrated, and only time will tell how accurate and influential these polls are.  But you only have to look at the effect on the news cycle of these polls and the ensuing speculation to realise that, whether media beat up, or genuine speculation, the destabilising effect certainly plays right into the hands of the rabid right.

So the question remains, if you don’t have a barge pole with which to shove Abbott over the line, will an opinion poll do the job?

Rupert Murdoch, practicing puppet master! (Photo credit: