A matter of opinion –
language and political influence

In a state of deceit, telling the truth is a revolutionary act. George Orwell

“A voters survival kit” Australian Liberal Party 2010Kevin Rudd is definitely not the man you want on the megaphone during an emergency and probably not even the best person to place the order at the drive through. As a communicator, Rudd suffers from a lethal mix of ailments: terminal prolixity, rampant hyperbole, a preference for technocratese, prone to outbreaks of neologismitis and an occasional resort to the colloquial that owes an unacknowledged debt to Bazza McKenzie. Don Watson could devote an entire chapter to Rudd in his next book and cruel wits have already compared him to a thinking person’s GW Bush. That so many in the electorate, including myself, did not detect this flaw in Kevin 07 says a great deal about how badly we wanted Howard gone. Nevertheless, one of the principal reasons Rudd is now in the sin bin was his failure to sustain a coherent conversation with the electorate, especially during times when significant policy developments needed to be explained. The other critical factor in his demise was his now widely acknowledged autocratic leadership style and his failure to consult his own party on key decisions.

The role played by opinion polls in the events of last week cannot be overlooked. On the Saturday after Gillard was sworn in, the Sydney Morning Herald’s headline announced, guess what, a swing to Labor in the polls. A number of commentators (notably Hartcher in the SMH; and Steketee and Rothwell in the Australian) have made the point that the Labor Right cabal of Shorten, Feeney and Arbib (with the suggestion that Richo was barely hidden behind a rather large arras) used ‘secret ALP poll results’ reportedly conducted by UMR Research to convince caucus that Rudd was a dud and to persuade Gillard to stand. Hartcher refers to the comparative ‘word clouds’ of Gillard (‘strong’, ‘capable’) and Rudd (‘arrogant’, ‘weak’) identified in the poll sample and how the technique is used, as one of a number of methods, by pollsters and their clients to gauge public sentiment.

Someone who knows a great deal about word clouds and voter sentiment is the American pollster and communications consultant Frank Luntz whose company The Word Doctors (don’t be fooled by the budget website, this is a very pointy outfit) is the preferred supplier of the stuff from which spin is made. Luntz is Rumpelstilskin’s straw merchant. His clients include the GOP and a parade of corporates and multinationals in constant need of persuasive material and keen advice on its deployment. Luntz is also a regular on Fox News. No surprises there. Luntz came to prominence under the tutelage of Republican stalwart Newt Gingrich and went on to advise the Bush administration on various matters, including the best way to approach the use of the language of ‘global warming’, which he famously rebadged as ‘climate change’.

Luntz’s favourite aphorism sums up his mode of operation: ‘It’s not what you say, it’s what people hear’.

I mention Luntz both for his entertainment value (check his videos, he is a cross between Jerry Springer and Michael Moore) and, more seriously, as an example of the way the ‘science’ of opinion polling is used not only to reflect public opinion, but to influence it. For some insights on how related techniques (known as ‘sentiment mining’) are applied to Web 2.0+ and social media check the latest Background Briefing.

In the lead up to Copenhagen, a coordinated global public relations campaign was orchestrated with the aim of discrediting climate change science and shifting public sentiment. The leak of an apparently contradictory IPCC email planted a seed of doubt in the right place at a critical moment and, in Australia, was seized upon by Minchin, Abetz, Joyce and Abbott as a trigger to begin the campaign to unseat Turnbull and to simultaneously attack the government. Great Big New Tax was introduced into the political lexicon. Rudd and Wong had not taken the time to refine their language when describing their climate change initiatives. However, there was a widespread sense in the electorate that something needed to be done even if we didn’t quite understand the issue. When the doubt crept in, the government’s lapse in focus opened up the opportunity for the populist scepticism of Joyce to combine with the Coalition’s carefully crafted Mr. Blah Blah Blah image of Rudd. And it worked. Rudd’s kitchen cabinet was spooked and the ETS was shelved. Abbott succeeded in wedging the government by using language that generated negative word clouds in the minds of voters in key marginal seats. And when you are defending your hold on power, these are the votes that matter most.

It is widely recognised that Rudd’s standing in the polls declined rapidly after he deferred the ETS legislation to 2012. All the major polls reported that voters moved away from both Labor and Rudd in favour of the Greens and, to a lesser extent, the Coalition. Rudd’s reasons for the policy shift were reasonable but poorly explained. Then something very strange happened. A public narrative escalated in which Rudd was no longer Tin Tin, no longer likeable. Now he was Mr. Cranky Pants Blah Blah Blah and his use-by date was judged to have expired by the Labor caucus.

Consumer and voter sentiment is measured in terms of emotion. How a particular product, person or service makes you feel. Frank Luntz uses language analysis technology in his focus group sessions. Participants are asked to listen carefully to every word and phrase in, say, a political debate, and to respond by turning a dial on a hand held device. The dial is calibrated on a scale of 0–100. Luntz describes the scale to participants this way: 100 says ‘I really love it’ and 0 says ‘I really hate it’, with 50 being neutral. Note the emotive descriptors. The software averages group responses and displays them on a moving graph, which is visible to the participants during the session. Thus we have: The Worm. Luntz uses the same technique when asking a group to assess words and phrases associated with the environment or financial reform or climate change legislation.

A number of factors are at work here. Firstly, there is the conversion of rational thought into reactive emotion. An assessment of policy initiatives, for example, requires a consideration of a range of highly complex, relative determinants. And yet a poll conducted in the way I have described requires the participant to reduce that complex thought process into a split second, gut reaction: a feeling.

Secondly, human beings are pack animals. We form our opinions as a part of a herd. When we see the majority of the herd heading in a particular direction our instinct is to follow, especially if the decision to do so is compressed into a microsecond. This results in the ‘bandwagon effect’ in which polling influences public opinion by driving people to the perceived majority view. Another term often used for an altogether different instinctual bias is the ‘underdog effect’, which compels politicians to stress the ‘difficulty of the task ahead’ around election time.

As Jeff Sparrow suggested in his Overland blog post on personality politics, it is ridiculous and even dangerous to assess the merits of political parties on the basis of the perceived personality traits of their leaders. It is equally absurd to believe that opinions formed on the basis of sentiment are the equal of informed, rational judgments. And yet, the predominance of polls in the political machinery is such that the nation’s leader can be deposed on information derived from techniques which shape public opinion as much, if not more, than they reflect it. More concerning still is the thought that sentiment polling could in future be used to drive ‘participatory democracy’. It is clear that opinion polling is effective in predicting human behaviour. My concern is that the techniques need to be better understood so that whether we are participating in a poll, reading the results of one or, more importantly, making decisions based on polls, we are fully aware of what we are doing and what is being done to us. The tight relationship between pollsters, political interests, business and the mass media makes it very difficult to untie this Gordian Knot.

Boris Kelly

Boris Kelly is a Sydney-based writer with an interest in theatre, literary fiction and politics. In 2009, he was the recipient of a Varuna Fellowship for work on his first novel.

More by Boris Kelly ›

Overland is a not-for-profit magazine with a proud history of supporting writers, and publishing ideas and voices often excluded from other places.

If you like this piece, or support Overland’s work in general, please subscribe or donate.

Related articles & Essays

Contribute to the conversation

  1. Great post, Boris. Your point about polling not only gauging public opinion but influencing it, is so rarely raised. There’s also the issue about the way polling (questions) can be framed to elicit certain responses with results used to serve political agendas. And while I fully support your comments about Rudd’s bureaucratic blah blah (who could not), the danger inherent in Gillard’s very articulate and polished performance, is that it won’t be so easy to identify that she too obfuscates the issues.

  2. Two good points, Trish.

    You wrote:
    “There’s also the issue about the way polling (questions) can be framed to elicit certain responses with results used to serve political agendas.”

    I’d like to hear some further analysis on this. Polling is integral to so many aspects of our lives yet the methodology receives very little informed coverage.
    For a snapshot of ‘sentiment mining’see http://www.youtube.com/watch?v=sIFYPQjYhv8

    You wrote:
    “..the danger inherent in Gillard’s very articulate and polished performance, is that it won’t be so easy to identify that she too obfuscates the issues.”

    This is evident in her neat explanation for the move on Rudd, that it was a matter of the government having “lost its way”. And yet the polls have moved to Gillard without a single substantive change in policy having been announced and only a minor tinkering in the cabinet reshuffle. What is the government if not the cabinet and its policies? To be fair, I think Gillard will be a better leader and communicator than Rudd but how that translates to policy and, more importantly, timely delivery (on which this government has been woeful), is yet to be seen.

  3. Trish, I should have said a snapshot of the scope of social media which, in turn, demonstrates the lure of sentiment mining for marketers and message makers using analytical language techniques. The Background Briefing episode on this is well worth a listen.

  4. Great post. I’ve been wondering about polling recently given the discrepancy between a number of the polls—particularly the Roy Morgan poll which seems to have different results than what has been popularly reported. (Not just the no honeymoon poll but also the poll previous to the removal of Rudd).

    The charting of the Roy Morgan polls against economic factors seem worthwhile but I find it interesting that they haven’t mapped a lot of this data against headlines or included a graph for all parties not just two party preferred. I wouldn’t mind see headlines before and after the polls are taken.

    In the media it appears that the polling is taken and assumptions are made so while a fair amount of support had gone to the Greens, it is hard to tell why this was specifically, mostly this is reported as due to the ETS, but could it also be refugees? And if it is the ETS why has there been a loss of support from the Greens back to Labor after Gillard becoming PM.

    Also if you read the last Roy Morgan poll something interesting is going on. Family First is unchanged, Independents are up 1.5%, the Greens fell by 3.5%, Labor fell by 2.5% and the Liberals went up by 4.5%. Suggesting that Greens support went to Labor and some Labor went to the Liberals. Or that some Greens support went to the Liberals. That being said the polls are quite volatile from week to week.

    Though given all that whenever I think of polling I always think of this scene from Yes (Prime) Minister:

    [To make this more frightening, the one time I went to the National Union of Students conference a number of the Young Liberals spent much of their time huddled around a laptop watching Yes Minister episodes]

Leave a Reply

Your email address will not be published. Required fields are marked *